As a kid, my conception of nanotechnology was this: a scaled down metal robot ship that could navigate the body and work at a molecular level. In the television show “The Magic School Bus”, the class in their titular vehicle often ventured into the human body in exactly this manner. This view of nanotechnology however is a misconception. Much of the work required in the body is performed at the atomic and molecular level, affecting individual atoms or groups of atoms.

Those interactions require incredibly small machinery and are dictated by the physical properties of how small particles interact at close proximity. In other words, the size, shape, and composition of the machine will directly affect its ability to act on some molecule or atom. For example, if some part of it has a positive charge due to specific atoms in its structure, this will affect how it interacts with other molecules and if it can even get inside the cell. To perform these functions, that machinery will not look like a robot, but will actually look more similar to the machinery that cells have already evolved to perform this function: enzymes, organelles, and other organic/inorganic compounds. In biotechnology: drugs, biologics, and now nanotechnology can be thought of as either manipulating your cells’ existing machinery, recreating and mass producing them, or synthesizing new types for biological use.

In this primer on nanotechnology, we are going to look at the field that is pioneering new types of nanoscale “machinery” and tools to interact with the biological components that make up your body. I’ll first discuss the basics of how to think about the field in terms of the science involved, move into the structure and composition of a nanoparticle, and then finish with various examples of nanotechnology used today and how those work.

Definitions:

Small Molecule Drugs:

• Chemically synthesized molecules that are designed to interact with specific biological pathways in the body.

• Some otherwise defined biologics like antibiotics are considered small molecule drugs because of their small size, relative simplicity to create, and major differences in their development processes from other biologics

• Examples include aspirin, antibiotics, and statins

Biologics:

• Substances that are derived from living organisms or their cells

• In some cases biologics can be whole cells like in the case of engineered T cells

• Other examples include hormones like insulin, antibodies, and vaccines

Nanotechnology:

• The creation of materials or substances like nanoparticles that can perform functions on the 1–100 nanometer level in cells — like delivering particles into cells or repairing damage. For reference a DNA double helix is around 2nm in diameter and a water molecule is .275 nm

• Examples include the nanoparticles that carry mRNA vaccines and other therapeutics into cells.

The Basics of Nanotechnology:

Nanotechnology in its real form looks a lot more like existing biological machinery than miniature robot ships. However, engineers have still created incredibly sophisticated nanomaterials and machinery. Generally, bionanotechnology consists of a core/scaffold, a linker group, and a biofunctional group. The core is oftentimes made of inorganic metals like gold, silver, iron oxide and silica but it can also consist of carbon structures or particles called quantum dots. Quantum dots are semiconducting compounds that have special optical and electronic properties useful for nanotechnology applications.

The linker group is determined by the core and the functional group chosen for the use case and serves to bind the biofunctional group to the core. Lastly, the biofunctional group is a molecule that is attached that gives the entire system biocompatibility. These could be ligands, DNA, proteins, and more. Their purpose can be to allow the system to interact with other targeted biological molecules in a body or sample, to allow the molecule to be more stable around biological molecules, or to introduce chemical effects like polarity, and more. In many bionanotechnology use cases, physical and chemical properties from both the core group and the biofunctional group are leveraged to obtain the desired outcome.

Properties at the Nanoscale

Understanding the physical and chemical effects at play on the nanoscale is vital to understanding and engineering effective nanotechnologies. One of the most important concepts is surface energy. Surface energy is the amount of energy needed to create or maintain an amount of area of surface of a particle. It arises as a result of the ratio of atoms that are at the surface rather than in “the bulk” or core of the particle. Due to geometry, the smaller a particle gets, the higher percent of its total number of atoms are on the surface. This results in very high surface energy for nanoparticles making them extremely susceptible to reactions in the environment, including aggregation amongst themselves. If you remember your high school chemistry, since the surface energy is very high for these atoms, interactions with the environment are likely to require less energy and thus create a net positive energy reaction — making them much more likely to occur.

Other very important properties include magnetism, polarity, and optical properties (surface plasmon resonance, fluorescence). All these properties are carefully considered and play important roles in component designs of nanoparticles.

Working with and Making Nanoparticles:

Various methods are employed to mitigate the aforementioned effects, especially surface energy during nanoparticle creation. The high surface energy of nanoparticles increases their tendency to clump up together, which is the last thing you want to happen when your goal is to create small particles. Nanoparticles synthesized in a bottom up fashion typically follow a three phase synthesis process: Atom Generation, Self-Nucleation, and Growth. During generation of atoms, the building blocks for the nanomaterial are created by breaking down precursor molecules via processes like chemical reduction. For example, gold nanoparticles are a very common choice to make up the core of a nanoparticle. These are generally created using HAuCl4 (tetrachloroauric acid) and a variety of reducing agents (chemicals that influence reactions by donating electrons). Choice of the reducing agent can have significant effects on the size and shape of the resulting gold nanoparticles because they influence things like speed of reaction, pH, and surrounding chemical byproducts.

The second phase, self-nucleation is when the resulting atoms, like gold in this case, start to combine to form the initial seeds of the nanoparticle. This provides the starting ground for the third phase, growth. This is the final stage of nanoparticle synthesis for at least the core component of the future nanoparticle. Growth is driven by a few processes like diffusion, coalescence, and Ostwald ripening. These processes describe the movement of atoms towards the nuclei by concentration gradient, merging of smaller particles to form larger ones, and the breaking apart of smaller particles into monomers that then attach themselves to larger particles respectively. Since this phase ultimately creates the core of the nanoparticle it’s important to implement strategies to ensure the right size and consistency. This is achieved by moderating starting concentration, temperature, pH, reducing agents, and addition of surfactants (molecules that bind and block additional growth in part or all of the seed). For example, temperature has a large influence on speed of reaction, lower temperatures will allow more gradual growth and can help ensure smaller and more consistent particles.

Other synthesis methods could include top-down synthesis by which a larger bulk of substance is used to create nanoparticles by means of “whittling away”. This can be done for example with a process called lithography that takes advantage of certain materials sensitivity to light. When light hits them they are susceptible to either dissolution or hardening. In these use cases, a photoresistent mask is partially placed over the substance so that the light interacts with only part of it. This is the method used to carve the extremely intricate patterns of circuits on our modern day computer chips that can contain well over 100 billion transistors (parts of circuits that turn it on/off).

Biofunctionalization:

In order for a nanoparticle to become a bionanoparticle, a functional group must be added that gives it biological properties. In order to get to the functional group, we must first return to the linker component of a bionanoparticle. The linker is the component that binds the core nanoparticle, like gold, to the biofunctional group. It consists of two parts: the surface binding group and the spacer.

The surface binding group is the component that binds to the surface of the core particle and is typically bound through a process like chemisorption or physio-sorption which involve strong covalent bonds or weak Van der Waal’s forces respectively. The choice comes down to a mixture of use case and core compatibility, physio-sorption supports reversible/dynamic processes versus the strong and permanent attachment of functional groups in chemisorption. Additionally, materials like gold are more amenable to chemisorption versus materials like carbon being more amenable to physio-sorption. For gold chemisorption, the common surface binder is a sulfur containing thiol that creates a very strong bond. This is then attached to a spacer molecule that creates a physical gap between the functional group and the core particle. In gold thiol nanoparticles, common spacers are hydrocarbon chains and polyethylene glycol. Polyethylene glycol particularly is known for its biocompatibility and is widely found in FDA approved nanoparticle formulations.

Similar to the selection process for surface binding groups, the selection of the functional group is dictated by the use case you are engineering for. There are two primary considerations: strength of bond required and compatibility with the targeted species. Strong bonds help to ensure that the interactions happen as consistently as possible, like in diagnostic tests. Weak bonds are useful for use cases that bind to a target molecule and release it on demand. The second consideration is self-evident, you need to use a functional group that is capable of binding to the target of interest in the body or solution.

Major Biomedical Use Cases:

In this section we’ll learn why this field of nanotechnology is a part of the broader umbrella of biotech. I’ll outline several use cases in medicine being used or explored today and their respective nanotechnology components. Some of these will follow different composition patterns than previously outlined due to the immensity of the field of nanoengineering and constant innovation therein.

Biosensors and Diagnostics

One of the most widely adopted current use cases for nanotechnology is in biosensors and diagnostic tools. Selective binding to targets of interest combined with leveraging unique chemical and physical properties of the nanomaterial provides an excellent environment for finding and/or manipulating biological materials in bodily tissue or fluid samples. Three prominent use cases are lateral flow assays (ex: pregnancy tests), nanostructured electrodes (ex: continuous glucose monitoring), and magnetic nanoparticles (ex: MRI contrast).

Lateral flow assays make use of the exact type of nanoparticle we’ve taken as an example thus far: gold-thiol nanoparticles. Lateral flow assays are used in test strip based diagnostics like pregnancy and antibody tests. Taking pregnancy tests as an example, they work by populating the sample side with gold nanoparticles functionalized with antibodies for human chorionic gonadotropin (a hormone produced during pregnancy). As the urine sample flows through the strip, the gold nanoparticles will bind to the hormone and will eventually meet the test and control lines we are familiar with on these types of tests. At the test line, there are immobilized antibodies for the same hormone. If the hormone is present in the urine and thus bound to the gold-antibody particle, it will form a sandwich complex with the immobilized antibodies like so: immobilized antibody — HCG hormone — gold-antibody conjugate. Since they are immobilized in a tight area, this creates a high density concentration of gold nanoparticles. Through a phenomenon exhibited by these particles called localized surface plasmon resonance, a red/pink hue is given off creating that visible line on the test. The particles at the control line meanwhile bind to excess gold nanoparticles regardless of hormone presence and takes advantage of the same phenomenon to produce its own line.

Nanostructured electrodes are electrodes that have been modified with nanostructures to help them perform a specific task. In the case of continuous glucose monitors, the electrode is modified with carbon nanotubes that in turn have immobilized glucose oxidase enzymes attached to them. The glucose oxidase enzymes break apart glucose and create gluconic acid and hydrogen peroxide. This hydrogen peroxide, being unstable, oxidizes at the electrode surface which acts as an electron acceptor. Because every 1 glucose molecule creates 1 hydrogen peroxide molecule and every 1 hydrogen peroxide molecule creates 2 electrons, we can calculate the amount of glucose in the interstitial fluid based on that electrical signal.

Lastly, magnetic nanoparticles are nanoparticles that take advantage of the inherent magnetic properties of the core metal component. MRI scans work by exciting the protons in water molecules and aligning them towards the magnetic field produced by the machine. The machine then produces a radiofrequency pulse that flips the proton alignment away from the magnetic field and causes them to spin. As the protons begin to re-orient to the magnetic field and stop spinning, this lag time is measured by T1 and T2 relaxation times respectively. Different tissues have different realignment times in both of these measurements and those lags are used by the machine to construct the image. Without contrast agents, whether you weight the image on T1 or T2 will affect what body tissues or fluids appear light or dark. Magnetic nanoparticles are used to affect these relaxation times and can be modified with targeted ligands so that they attach to tissues like tumors or other cell types. When these nanoparticles clump at a specific location in the body, they cause the resulting MRI image to darken at that spot.

Drug Delivery Tools

Another both widely adopted and growing area for bionanotechnology is in drug delivery systems. The most famous of these is the lipid nanoparticles that were developed and used for use to deliver mRNA vaccines during the Covid-19 pandemic. Additionally, a lot of research and clinical trials are ongoing for exploration into other types of delivery vehicles like the versatile gold nanoparticle which shows promise for targeted cancer therapy.

The formation of lipid nanoparticles relies on the phenomenon of self-assembly in nanotechnology. This is the natural process by which molecules bond, change orientation, and otherwise assemble together based on their chemical and physical properties. This is how things like protein folding work in your own body. In the case of lipid nanoparticles, self assembly occurs due to the interactions between the different lipids and mRNA particles. To create these lipid carriers, the mRNA is mixed with phospholipids, cholesterol, ionized lipids, and PEGylated lipids (polyethylene glycol mentioned earlier). Because the ionized lipids have a positive charge they are attracted naturally to the negative charge on the RNA and orient around them. At the same time, both these lipids and the phospholipids, cholesterol, and PEGylated lipids are interacting to naturally form spheres due to their hydrophilic heads and hydrophobic tails — the same mechanism as your cell membranes. The lipid nanoparticle is crucial to allowing the mRNA to avoid being broken down in the body and for enabling it to enter the cell. Additionally, the PEGylated lipids (lipids with PEG attached to them) produce branches extending outward that help limit interactions with the immune cells.

Other Use Cases Being Explored:

• Tissue scaffolds — nanostructures to provide a scaffold for new tissue like bone with embedded enzymes and growth factors to promote regeneration.

• Antimicrobials — designing nanoparticles as synthetic antibacterial agents

• Theranostics — selectively attaching nanoparticles that are then selectively activated to release a drug being carried or to otherwise harm the target (examples include nanoparticles activated by light or magnetic fields after binding to a tumor)

• DNA hydrogels — these hydrogels of DNA and transcription factors can produce functional proteins in the absence of a living cell. They can also be selectively activated.

The field of nanotechnology is extremely deep and we’ve just scratched the surface in this post. There is a vast literature on other types of nanoparticles and their synthesis, like quantum dots, fullerene, and nanodiamonds. Additionally, I’ve only briefly touched on the complex physical and chemical properties involved in nanoparticle engineering. While I’ve described some of the most mature uses of bionanotechnology there is a large variety of fascinating uses that I highly encourage you to explore. My favorite is DNA origami — complex structures built out of DNA using the predictable binding of complementary base pairing.

Subscribe now

Introduction:

In Post World War I London a Scottish scientist named Alexander Fleming was immersed in research to discover better ways to fight infections. Inspired by his observations of infections often progressing despite administration of antiseptics while serving in the Army Medical Corps during the war, he was inspired to research antiseptics afterwards. His first discovery was that of the lysozyme enzyme that exhibited modest antibacterial properties in the petri dishes seeded with his own mucus he gathered during a cold infection. While a great first discovery, it lamentably had only a small effect on a subset of bacteria.

In 1928, he began experimenting on the staphylococcal bacteria. Notoriously disorganized, one of these petri dishes was left uncovered near an open window, something that became a serendipitous miracle for mankind. When he discovered that this petri dish compared to others did not produce bacterial colonies, he investigated the dish and found it contaminated with mold spores from the Penicillium fungus. With further investigation, he isolated the effect to a “juice” secreted by the mold which he dubbed penicillin. Penicillin, he found, showed remarkable antibiotic properties against all gram+ bacteria (bacteria lacking an outer membrane) which are responsible for illnesses like strep throat, scarlett fever, and pneumonia. This stroke of luck kicked off an antibiotic golden age of mass production and discovery over the following decades. Some examples of integral antibiotics today:

• Amoxicillin - 1972

• Parenteral Vancomycin - 1953

• Cefazolin - 1971

• Piperacillin-Tazobactum (pip-tazo, Zosyn) - 1993

How Antibiotics Work:

Generally, antibiotics work by targeting and disrupting essential cellular pathways within bacteria. These targets can include disruption of the integrity of the cell walls or membranes, protein production, or DNA/RNA replication/transcription. In the case of all of the above antibiotics listed, the primary target is in cell wall synthesis. Cell wall integrity is essential to a bacteria’s survival but they are not always capable of being targeted by a given antibiotic. The most common reason is the presence or lack of an outer membrane. As we touched on briefly, gram+ bacteria lack an outer membrane while gram- bacteria have one. This limits effectiveness of antibiotics on gram- bacteria based on whether or not they can make it through the outer membrane.

• Amoxicillin - broad coverage (gram+ and some gram-)

• Parenteral Vancomycin - gram+ only

• Pip-tazo (Zosyn) - very broad (gram+ and gram-)

• Cefazolin - mostly gram+

How Antibiotics are Used:

The targeting capability of a given antibiotic is important to understanding its importance and use in a clinical setting. Generally speaking, many clinical protocols like the sepsis protocol (sepsis is a bacterial infection in the blood stream), require broad coverage antibiotics to be given within 1 hour of a sepsis diagnosis. In the case of our listed antibiotics, Zosyn (brand name of pip-tazo) is a core part of the sepsis bundle and serves an important role in initiating broad coverage immediately. For a dangerous condition like sepsis, the goal is to immediately start counteracting the bacteria even though you don’t know exactly what type it is yet. If MRSA (methicillin resistant staphylococcus aureus) is suspected, IV (parenteral) vancomycin may be included in this initial stage. Meanwhile, samples will be sent to the hospital lab for speciation of the bacteria.

When cultures return, Zosyn/pip-tazo will be tapered off, and depending on the findings, new antibiotics will be administered to specifically target the offending bacteria. For example, if findings show that the bacteria is methicillin sensitive, Cefazolin may be given during this refinement stage. This stage is crucial to making sure that the patient’s microbiome is as minimally affected by antibiotic treatment as possible. Prolonged use of powerful broad coverage antibiotics like Zosyn increase the odds of developing antibiotic resistance, conditions like clostridium difficile, and generally disrupting the patient microbiome – something that we increasingly see carries with it effects all over the body. Amoxicillin in contrast to these other antibiotics, is generally given in outpatient settings as the first-line treatment for most of those uncomplicated infections like strep throat and urinary tract infections.

• Amoxicillin - most popular outpatient setting antibiotic

• Vancomycin & Pip-tazo - two of the most popular and widely used in the critical care setting (ICU)

• Cefazolin - most popular antibiotic for surgical prophylaxis (prepping surgical patient prior to surgery to minimize likelihood of bacterial infection from surgery)

Antibiotics revolutionized healthcare for human civilization, helping to cut mortality from bacterial infections from the number one cause of death in the US to the 8th. They’ve also enabled much of our modern medicine around surgery and other invasive procedures that used to carry high risk of infection. Medicine as we know it would not be possible without these drugs. In terms of critical goods, it’s hard to make a better case than antibiotics – so how are they made and how do we supply our healthcare system with them?

Antibiotic Production Process:

Antibiotics, depending on the type, follow a few different production paths: natural-product, semi-synthetic, or fully synthetic. Natural-product antibiotics are produced according to biologics manufacturing principles: namely the cultivation of cells that produce the desired substance (either naturally or post-gene-modification) followed by the harvesting of their antibiotic product when a sufficient quantity has been produced. One example is penicillin from the penicillium fungus. For an end-to-end breakdown of the biologics manufacturing process I’d recommend my previous post on the subject.

For semi-synthetic antibiotics, the first part of the manufacturing process mirrors that of natural-products. After a biological core has been created and purified from the biologics manufacturing steps, the next stages are devoted to chemically altering the substance with desired side chains. For example, amoxicillin leaves the biologics process with a core of 6-APA (6-Aminopenicillanic acid), afterwards during the synthetic modification stage, it is modified with a side chain of p-Hydroxy-D-phenylglycine in order to improve oral tolerability and increase effectiveness against certain bacterial types.

Fully synthetic antibiotics skip the biologic manufacturing process completely and rely purely on synthetic organic chemistry reactions. This type of pharmaceutical manufacturing parallels what is commonly called “small-molecule-drug manufacturing” and I outline that process with Aspirin as an example in a previous blog post. This type of manufacturing is generally limited to smaller and less complex molecules as the number of independent reaction steps can greatly influence the complexity and cost of the manufacturing process. Fully synthetic antibiotics need to be those that can reliably produce tonne-scale active pharmaceutical ingredient (API).

As opposed to highly-tuned biological processes isolated in a cell, fully synthetic processes have to contend with isolation of intermediates, protecting them from undesired effects, purifying them, and more. In biology, this entire process is managed within the cell. A cell can run dozens of chemical steps in parallel, regenerate its own catalysts, manage its own relative ratios of reactants (stoichiometry) and isolate different materials in the “supply chain” of the process within cellular compartments rather than in different reactor equipment required in a fully synthetic process.

Active Pharmaceutical Ingredients and Finished Dosage Forms:

Active pharmaceutical ingredient is the finished biologically active product of the drug manufacturing process prior to being converted into a “finished dosage form” or FDF. It is important to note that many people use API in a way that suggests it is a sort of raw material. However, the API is a finished product in its own right. In an analogy with vehicle construction, the API is like the machined car parts while the finishing stage is like assembling the car itself. In many ways the fabrication of the car parts encompasses a much higher amount of the work gone into the finished car. This is especially important to understand when talking about API supply chains as treating APIs like a raw material overlooks the complexity of their manufacture.

In the case of amoxicillin that we touched on briefly, the API is the fully purified substance of 6-APA plus the p-Hydroxy-D-phenylglycine side chain. In many cases, this API is then shipped to another site that performs the formulation and fill finish steps that for amoxicillin adds excipients (chemicals that make the API safer, more stable, and biocompatible). Other examples in this phase include diluting the API so that it is the right potency and coating it with materials that aid in swallowing or the timed release of the substance in the human body. When a FDF finishes this stage and passes quality checks it can then be shipped to a clinical site for use. Once there, depending on the substance, it might require a small amount of work done by the hospital pharmacy team to prepare it for clinical use like mixing dry product with water/saline for administration via IV bags.

Antibiotic Supply Chains:

Now that we’ve touched briefly on the general manufacturing process I want to turn to a supply chain mapping of two of the most crucial antibiotics for US healthcare: Vancomycin and Piperacillin-Tazobactum (aka pip-tazo, brand name Zosyn). These are pillars of infectious disease care in the US serving as front line cornerstones in the treatment of MRSA and sepsis respectively.

The Vancomycin Supply Chain:

Vancomycin, the first line treatment for MRSA and C. difficile, is a fully biologically produced compound. That means that the active pharmaceutical ingredient is produced by living cells within a fermentation bioreactor. Key inputs here are starches for carbon, soybean flour/meal for nitrogen, and inorganic salts like sodium nitrate/magnesium sulfate. The bacterium strain used to produce vancomycin is Amycolatopsis orientalis.

In brief, the vancomycin API is produced by microbial fermentation in large reactors (the vancomycin is secreted into the cell broth so no cell lysis step is needed), goes through a centrifuge to separate cellular and other materials from the vancomycin, is run through ion-exchange chromatography to purify it and ultrafiltration to concentrate and exchange buffers, lastly the vancomycin is precipitated, dried, and milled into powder producing the final API.

The vancomycin API (product produced during manufacture) is primarily done in China and India with over 60% of API production occurring in China. The formulation stage consisting of taking the API and making it ready for use often occurs in India for the US market. In Europe, the company Xellia produces both the API and the FDF via in-house formulation. The US does comparatively very little API production or formulation for vancomycin, primarily outsourcing to India or Europe. Notably, the US company Alpharma is involved in a joint venture with the Chinese company Hisun to expand vancomycin API production further in China – estimated to double supply by 2026.

The vancomycin supply chain for the US is heavily reliant on partners in China and India with only more reliance projected as a result of the Chinese-American joint venture. Vancomycin has been one of several drugs that has been in chronic intermittent shortages for around the past decade, with an active shortage declared by the American Society of Health-System Pharmacists (ASHP) as of publishing this piece (Jan 2026). It is possible that shortages may be remedied by the aforementioned supply expansion, but more and more attention is being paid to API manufacturing in general as the US seeks to establish more critical manufacturing independence.

The Pip-Tazo Supply Chain:

Pip-Tazo is a hybrid antibiotic in two senses: it is a hybrid between piperacillin and tazobactum, and is semi-synthetic involving both biologic and chemical synthesis for the piperacillin component. Piperacillin is a penicillin derived antibiotic, starting first with a fermentation step in order to create penicillin G. Penicillin G is then chemically modified to create 6-APA: the primary chemical precursor for both the piperacillin and tazobactum components. The remaining steps for both piperacillin and tazobactum are all chemical synthesis for the API manufacturing stage. It is not until the FDF stage that the two are mixed at clinical ratios into the end Pip-tazo product.

6-APA is the critical component to call out in a supply chain discussion for Pip-tazo and antibiotics in general. China has 5 out of the 7 plants in the world that makes 6-APA which is critical not only for pip-tazo but for all penicillin derived antibiotics including amoxicilin, ampicilin, methicilin and more. Downstream of 6-APA, to the piperacillin API, it is even more concentrated. In 2016, a global pip-tazo shortage and health crisis was triggered by an explosion at the sole plant in the world that could produce piperacillin which was located in China.

This concentration was driven by supply failures by Pfizer/Hospira and a failed Pfizer/Chinese joint venture in China. Throughout the 2000’s pip-tazo was in chronic shortages due to Pfizer’s failure maintaining supply, thought to be due to a decline in revenues with the emergence of generics. In the mid 2010’s, with struggles to profitably manufacture pip-tazo API, Pfizer attempted a joint venture with the Chinese manufacturer Hisun. This joint venture ultimately failed, leaving Qilu pharmaceuticals in China as the sole manufacturer. In 2016, a pressure vessel exploded at the plant and halted the entire global supply of pip-tazo.

Since that crisis in 2016, the pip-tazo supply chain has diversified, but is still mainly centered in China and India. The primary players are Qilu Pharmaceuticals and Shandong Ruiying in China and Aurobindo (Eugia) in India. In contrast to Vancomycin however, the FDF stage for pip-tazo in the western market is dominated by Europe and North America (Pfizer and Novartis in Europe, B. Braun, and Fresenius Kabi in the US). This is attributed to higher complexity and stricter quality requirements for the FDF stage of pip-tazo.

Securing the Supply Chain:

The supply shock of pip-tazo due to the explosion at the Qilu plant in 2016 is a valuable case study as to how fragile supply chains can become as specialized manufacturing becomes concentrated into one company or country. We saw that financial difficulties forced the brand name manufacturer Pfizer to first outsource manufacture, and then shutter. This aligned with the end of patent for Zosyn and emergence of generics. No longer having a monopoly on pip-tazo, Pfizer appears to have failed to retain financial viability in its production e as prices fell. This prompts several questions in determining ultimate causes: did monopoly cause Pfizer to get lazy with optimizing manufacture or did it merely decide that energies were better directed towards more profitable endeavors? If production could still be profitable but ending it was due to opportunity cost, why did generics API production concentrate into China and particularly one company? Presumably, if generics can be manufactured profitably in the West even at lower profit margins, there is nonetheless a market opportunity for those local manufacturers.

There is another possibility: the drug itself is not profitable to manufacture without market power. With large scale industrial policy backstopping many industries across China, it is feasible that an unprofitable drug could be picked up in the short term by Chinese manufacturers until a monopoly secures market power. Other factors at play make outsourcing to countries like China for API manufacture attractive, but it is interesting that in this case the outsourcing failed and it was an independent Chinese generic manufacturer that won. The general trend of concentration of API manufacturing across drug classes within China and India suggests structural barriers to re-shoring API production in the West.

Today, western API production consists of mainly expensive specialty drugs, which gives credence to the idea that western manufacturers require large top line revenues to offset more expensive production. Or to the theory that they instead actively choose to only produce drugs that benefit from large profit margins enabled by patent protection. In a blow to the second theory, even branded API manufacture is increasingly concentrating in Asia. If the US government is serious about its continued reference to reshoring API production, both policy makers and entrepreneurial minds in the pharmaceutical industry would do well to understand exactly what gives the Asian manufacturers competitive advantage and which can be overcome by policy, which can be overcome by smart technological or business innovations, and which we may have to settle for “friend-shoring”.

Subscribe now

References:

https://pmc.ncbi.nlm.nih.gov/articles/PMC4520913/

https://www.ashp.org/drug-shortages/current-shortages/drug-shortage-detail.aspx?id=25&loginreturnUrl=SSOCheckOnly

https://www.cidrap.umn.edu/resilient-drug-supply/some-critical-drugs-have-been-shortage-more-8-years

https://www.fieldfisher.com/fr/locations/france/insights/navigating-new-risks-how-china%E2%80%99s-anti-espionage-la?utm_source=chatgpt.com

https://pmc.ncbi.nlm.nih.gov/articles/PMC11616744/#:~:text=capitalism%2C%20China%2C%20Pfizer-,Background,Perry%20&%20Markham%2C%201999).

https://www.nature.com/articles/s41599-024-03759-y

“Biology is the most powerful technology ever created. DNA is software, protein are hardware, cells are factories.” — Arvind Gupta

Coming off our primer on small molecule drugs, we learned that those medications are chemically synthesized into compounds that have some effect on our biological machinery when introduced to the body. As we come to biologics, we will learn that not only are we affecting our own biological machinery when they are introduced, but we are directly harnessing biological machinery for their production. Biologics, or biopharmaceuticals, are biological compounds that are derived from, produced by, or otherwise made of cells.

The creation of biologics is incredibly recent in comparison to the history of small molecule drugs. Advances in cellular biology, engineering, biochemistry, imaging and more were vital pre-requisites to biologics development. We’ll see in the history and the present of biologics manufacture that the harnessing of cellular machinery introduces more challenges than traditional drug manufacturing. Due to the size and complexity of biological compounds, the use of chemical synthesis is currently impossible. Since artificial chemical synthesis is off the table, we must rely on the source we already know works to create these compounds: living cells.

In the rest of this post we’ll:

• Review the definitions of the molecules in this series

• Dive into the history of insulin and insulin manufacture as the flagship biologic

• Outline how biologics manufacture occurs today

• Discuss new areas of innovation occurring in the industry

Definitions:

Small Molecule Drugs:

• chemically synthesized molecules that are designed to interact with specific biological pathways in the body.

• Some otherwise defined biologics like certain antibiotics are considered small molecule drugs because of their small size, relative simplicity to create, and major differences in their development processes from other biologics

• Examples include aspirin, certain antibiotics, and statins

Biologics:

• Substances that are derived from living organisms or their cells

• In some cases biologics can be whole cells like in the case of engineered T cells

• Other examples include hormones like insulin, antibodies, and vaccines

Nanotechnology:

• The creation of materials or substances like nanoparticles that can perform functions on the 1–100 nanometer level in cells — like delivering particles into cells or repairing damage. For reference a DNA double helix is around 2nm in diameter and a water molecule is .275 nm

• Examples include the nanoparticles that carry mRNA vaccines and other therapeutics into cells.

Insulin and Diabetes

Insulin aside from being the flagship use case of biologics is perhaps one of the most known pharmaceuticals today for its use in treating Diabetes Mellitus Type I and II. In normal functioning humans, the hormone insulin is secreted by the pancreas as a response to elevated glucose (sugar) levels in the blood in order to prompt changes in cells to counteract this elevation. The primary response is by triggering glucose uptake by muscle cells which account for 70–80% of insulin stimulated glucose uptake in the body. It does this by binding to a receptor on the muscle cell surface and activating a signal cascade in the cell that moves GLUT4 transporter proteins to the surface of the cell. When GLUT4 is exposed at the surface, it allows glucose to enter the cell.

In type I diabetes, the pancreas cannot produce insulin because the body’s immune system mistakenly kills the pancreas’s beta cells that produce it. In type II diabetes, the body’s cells have trouble responding to the signal of insulin to trigger glucose uptake due to several metabolic derangements in the body. These include increased presence of free fatty acids and inflammatory cytokines caused by too much fat tissue and mitochondrial dysfunction that ultimately inhibits and interferes with parts of the insulin signaling pathway. In serious cases of type II diabetes, insulin injections are required to override these inhibitory phenomena in order to lower blood sugar.

The primary cases of diabetes in the historical record relate to Type I Diabetes due to the primary causes of Type II Diabetes being a more modern phenomenon. Diabetes was initially noticed as a condition that caused excessive urination, thirst, and dehydration. The word diabetes was coined by the Greek physician Aretaeus of Cappadocia deriving from the Greek word for “siphon” referring to this loss of water. Later, the latin word mellitus, meaning “sugar sweet” was appended to diabetes by the English physician Thomas Willis. Although an odd diagnostic referring to the taste of the urine, it was an important revelation that sugar was appearing in the urine of these patients. This occurs in both Type I and II diabetics as the body will try and remove excess blood sugar by excreting it out in the urine. However, due to the process of osmosis, the increased concentration of glucose in the urine also pulls in excess water to balance out this high concentration. This is what causes the diabetic hallmarks of extreme dehydration and thirst, excessive urination, and presence of glucose in the urine.

For the vast majority of human history, there was no treatment for this condition and having it was a death sentence. This was made even more sad because type I diabetes usually arises in children and young adults. The prognosis for these patients was a gradual wasting away caused by the body’s inability to extract the sugar energy from the blood, requiring the breakdown of the body to fuel itself. This breakdown of alternatives like fatty acids causes the build up of acidic ketone bodies (Diabetic Ketoacidosis) damaging organs and leading to coma and death. Prior to insulin, the standard treatment for diabetes was a starvation diet extremely low in carbohydrates that allowed diabetics to live slightly longer than otherwise.

The History of Insulin:

The first step in solving any problem is understanding it. For diabetes this means understanding how and why it arises. The first breakthrough came in 1889 via an experiment by two German researchers: Okar Minkowski and Joseph von Mering. In this experiment, they removed the pancreas from dogs and found that dogs without a pancreas developed the same symptoms as diabetics. In 1921, insulin was finally isolated from the surgically removed pancreas of a dog and proven to reverse diabetic symptoms when re-injected into the dog. The successful isolation of insulin from animal pancreas and then the first successful treatment on a 14 year old diabetic patient named Leonard Thompson a year later sparked a huge interest in insulin production. In the US and Denmark within the companies of Eli Lilly and what became Novo Nordisk respectively, mass production of insulin took off to treat what was before a death sentence.

The issue however was that the only way that was known to “create” insulin was the same as what was used in the first experiment: grinding up an animal pancreas and then isolating the insulin from the resulting mush. The main difference in the mass production process was that now instead of using dog pancreas they used the leftovers from cows and pigs after slaughter. This crude way of extracting insulin was extremely inefficient. In fact it took 1000’s of cow and pig pancreas to recover enough insulin to treat one diabetic for one year. This put a hard cap on supply and made insulin a scarce commodity.

In 1978 however, the paradigm for insulin production and the entire field of biotechnology was changed forever. That year, scientists at the company Genentech in partnership with Eli Lilly successfully inserted the gene for human insulin into E. coli bacteria using recombinant DNA technology. The ability to add genes coding for a specific biological molecule to a cell that can be cultivated to a large colony and produce a large amount of product is what birthed the entire biologics industry. This is the paradigm that we fundamentally still operate in today.

Subscribe now

Biologics Manufacturing: Insulin Case Study

Biologics manufacturing is done by living cells because the processes to construct these large and complex molecules are far too intricate to be done by the relatively simple chemical reactions that make up small molecule drug creation. Thus, it relies on the already configured nano machinery of cells to ensure that the product is properly built. The reliance on living cells however, introduces much complexity into the manufacturing process. In this section we will go through each stage of the biopharmaceutical manufacturing process, explain the goals and constraints, and compare/contrast the modern and historical process.

Stage 1: Upstream Operations

Upstream operations refers to everything before product harvest. This means cell line engineering and cell cultivation. In modern biologics manufacturing, manufacturers maintain specifically engineered cell lines for certain products. These cell lines are called master cell banks (MCB) and are cryogenically frozen to maintain their state. It is critically important to maintain them unaltered because biological systems mutate over time as DNA replication errors compound. The cells that ultimately become producers of the biologic are formed from a working cell bank (WCB) that is descended from the master cell bank. The types of cells chosen depend on multiple factors like how quick they replicate, if they secrete the product or not, and how expensive they are to cultivate. For insulin, genetically modified E. coli cells are the common cell type used but many other cells are used in biologics manufacture. Some examples are yeast cells (S. cerevisiaie), mammalian (chinese hamster ovary), and human (human embryonic kidney).

Once a WCB sample is selected, it needs to be scaled to around 10% of the bioreactor working volume to optimize for the cellular growth curve during fermentation. Since a cell bank vial initially contains around 1mL it must be run through a “seed train” of progressively larger bioreactors until it is the proper volume to inoculate the production bioreactor. If the bioreactor is 20,000L then the cell culture that ultimately starts the batch will be 2,000L.

The bioreactor itself has two requirements: enable aseptic operations and provide the optimal environment for cellular growth and product formation. This has implications for the design of the bioreactor. Traditional bioreactors are stainless steel because the material does not leach into the cell medium, it is easy to clean, and it will not degrade from the harsh cleaning chemicals or the temperature extremes. There is also an increasing interest in single-use bioreactors made of polymeric materials that eliminate the need for cleaning procedures. In the case of traditional re-usable reactors the design is such that the cleaning procedures can most effectively remove any and all contamination from both the prior cell batch and the cleaning chemicals themselves. Other fascinating features of bioreactors today is their ability to continuously monitor the medium environment and feed the cells past what the original medium would have allowed them to grow and mixers that will ensure that extra added nutrients are distributed to all cells. The mixing blades are meticulously designed to ensure an even distribution without damaging the cells and different designs are better for tougher or more fragile cells.

When cells have exhausted their medium of nutrients or a certain amount of time has passed, this triggers the end of this phase. Remember that we don’t want cells living for too long because genetic mutations might occur that alter the product.

Historical Process:

This phase in the early days of insulin was occupied with procuring and grinding up the pancreas of livestock

Stage 2: Cell Lysis & Product Harvest

Cell lysis is a required phase for all biologics that are produced as intracellular products. Insulin for example is produced as an intracellular product typically within E. coli cells while monoclonal antibodies produced from something like chinese hamster ovary cells are secreted into the cell culture medium. Secretion of product helps to simplify the entire process because it removes this step and makes filtration and purification steps easier. Because insulin is retained within the bacterial cells, they need to go through a process to split the cells and release it.

There are a variety of cell lysis techniques that depend on both the cell type used and the product being recovered. The major categories of cell lysis are mechanical and non-mechanical. Mechanical cell lysis aims to use physical force to disrupt cell structural integrity and are used more frequently in animal cells since they lack rigid cell walls like bacteria and yeast. Common methods for mechanical lysis include high-pressure homogenization where cells are forced through a small orifice at high pressure and bead milling where cells are ruptured by shaking them with glass or stainless steel beads. Non-mechanical cell lysis relies on chemical effects and is frequently used for bacteria. Examples include chemical and enzymatic treatment that break down cell walls and membranes and osmotic shock treatment which forces water into cells causing them to break. In the case of insulin production, E. coli cells are treated with enzymatic treatment (lysozyme) to break down their cell walls and chemical treatment with a surfactant like Triton X-100 to break down cell membranes.

Historical Process:

Since the pancreas cells were animal cells the mechanical grinding up of the tissue sufficed for product harvest

Stage 3: Filtration

The filtration or clarification stage is focused on the separation of liquid from solid components in the mixture thus far. This is important to remove unwanted materials because the purification steps require absence of solid particles. In biologics manufacturing there are 3 main techniques for separating solid and liquid: centrifugation, depth filtration, and tangential flow filtration. These are often either used together or by filtration techniques alone.

Centrifugation is aptly named after the centrifugal force that it harnesses to settle denser matter farther away from the center of rotation than less dense matter. In the case of a centrifuge, that means along the bottoms and sides of the centrifuge vessel. Centrifugation however, while effective at removing much of the large particulate in the slurry, does not remove enough to be used on its own. Depth filtration is typically used after centrifugation where the intermediate mixture is fed into filter modules using a filter of cellulose fibers and polymeric resin that adds a positive charge to the membrane. The liquid is pulled against the filter using a pressure differential and particulates that are either too big for the pores or chemically attracted to the membrane are captured. In cases where depth filtration is used in place of centrifugation, it is common to take more than one pass through.

Tangential flow filtration (TFF) can also be used in place of the previous two and refers to a setup where the mixture travels through a membranous tube so that the flow of liquid is parallel to the membrane rather than directly against it in depth filtration. Pressure differentials encourage liquid to cross the membrane but because it is running parallel, it reduces solid build up in the membrane pores that prevents or slows down liquid perfusion. TFF is often used to prepare the product intermediate by concentrating it or exchanging the chemical buffer that it is in prior to purification steps.

In insulin manufacture today centrifugation is the first step after cell lysis because in this case, insulin itself is actually created initially within the bacteria as an insoluble inclusion body that is denser than much of the rest of the slurry. This means that after centrifugation, it is the solid insulin inclusion bodies that get extracted and move on in the process. After these inclusion bodies are extracted, they are solubilized and refolded in a buffer containing denaturants and other chemicals. From there filtration may occur to further clarify the solution, concentrate the insulin, or exchange the buffer before purification.

Historical Process

Insulin from pancreatic cells was already liquid so centrifugation operated in the reverse in that the liquid was retained and the solid cell and tissue mass was discarded. In contrast with today’s advanced cellulose and polymeric filters and pressurized systems, cloth and paper filters were used to separate more solid particles from the liquid by pouring it over the filter and letting gravity pull it through.

Stage 4: Purification

Biologics are unique drugs not only because they involve living cells, but also because of the way they are administered. Because of the chemical composition of these compounds, they cannot be taken orally or else they can be broken down or deactivated in the digestive system. Instead, these drugs need to be taken parenterally, meaning by infusion/injection. Because of this direct injection into the body, biologics have very stringent quality and safety standards. The purification stage is vital to ensuring these requirements.

After the clarification and filtration stage, the resulting intermediate is free of insoluble impurities. While a good first start, there are still many soluble impurities that have made it through and could seriously impact product safety and quality. Some examples are host cell proteins, endotoxins, viruses, and residual chemicals from upstream processes. Removing these is the purpose of the purification stage. Chromatography is the most popular technique for this stage and will thus be the focus here.

Chromatography is a chemical separation technique that was first used to separate plant pigments — where the technique got its name (chroma meaning color). It operates on the fact that through the uniqueness of the chemical structure of different molecules in a fluid mixture, they will pass through a given “filter” at different speeds. This is determined by their molecular interactions with the filter. The types of interactions needed will dictate the type of chromatography being used and could be based on chemical binding affinity, electrostatic binding based on charge of the molecule and filter, or size exclusion based on the size of the molecule. In chromatography, the mixture including the product is referred to as the mobile phase, and the “filter” that is packed within the column that the mixture flows through is called the stationary phase, or the resin.

It is very common to go through multiple passes of different types of chromatography to progressively filter out various types of impurities. In insulin for example, the process could go through three different types. Starting with ion exchange chromatography, when the mixture is poured through the resin, insulin and other similarly charged molecules will be electrostatically bound to the resin while other molecules considered impurities without charge affinity will flow through freely. Then a new chemical is poured in that changes the electrostatic environment of the column so that the insulin unbinds and flows through. After that, affinity chromatography will be used where the resin contains insulin specific ligands for the insulin to bind to while other non-specific impurities are washed away. Like the previous step, a new chemical will be introduced to break the bonds and allow for the insulin to be removed purely. Lastly, reverse phase chromatography may be used based on hydrophobic interactions. Insulin, containing hydrophobic regions will bind via hydrophobic interactions with the stationary phase while mis-folded insulin variants will wash through.

The end result of this process is both very pure in that it is free from non-insulin impurities but also of high quality in that it is free of improperly constructed insulin.

Historical Process:

This phase historically was a little different and relied more on crystallization of insulin. The steps include first precipitating insulin out using alcohol, then running it through ion-exchange chromatography (based on charge), and then again crystalizing it using zinc and washing the crystals. In early production of insulin based on animal insulin, it was too common by today’s standards to have immune reactions to animal host cell material not successfully removed.

Stage 5: Formulation

While the resulting product intermediate from the last phase of chromatography is very pure, it is also very likely to be at a different concentration and within a different buffer than is required for the end drug product. Formulation is the phase where the drug substance is fully prepped to be converted to drug product when ready. One technique you may probably already guess is coming is tangential flow filtration in order to adjust concentration and exchange buffer. This is indeed used but in a slight variation called ultrafiltration. The main difference is that the membrane is setup to retain the product while the solvent filters through, making it ideal for buffer and concentration adjustment.

It is important to note that the formulation phase ends the production stage of drug substance rather than drug product. The difference being that the drug product is that which is fully ready for medical use. It is common for the drug substance produced at the end of the formulation stage to include excipients that modify drug behavior as well as make it more stable for storage or transport. In the case of insulin, these may be preservatives to prevent microbial growth as well as modifiers that slow down insulin absorption in the body. This finished drug substance is bulk filled into a container for frozen or refrigerated storage. When the product is ready to be shipped, it is thawed, pooled, additionally compounded with excipients if required, and then filled into the final drug product container like a vial or syringe.

Historical process:

After the final purification step of crystallization, the crystals needed to be dissolved and compounded with excipients. This process contains many similarities with today in the types of excipients involved to stabilize and preserve it. The main differences were that historical insulin formulations did not have the same modifications that changed the absorption profile of insulin.

Innovations in Biologics Manufacture

As we see from the analysis of the previous case study, biologics manufacture has made enormous advancements in the technology we have brought to bear on it. Advancements in genetic engineering first allowed us to hijack cells like bacteria to produce human proteins for us rather than relying on harvesting small amounts from animals. Further advancements in genetic engineering have opened the doors both to more efficient target protein production but also new types of therapeutics like CAR T-cell therapy. This therapy modifies a patient’s own immune T-cells to attack a specific target. It has shown to be highly effective against blood cancers and new data shows a lot of promise for curing autoimmune diseases as well.

Advancements in machine technology and biochemistry have also allowed production to scale and reach incredible levels of purity and safety. Emerging in the field are new technology trends around:

• Single-use equipment

This allows intense development of cleaning procedures to be bypassed because equipment is discarded after the batch.

• Continuous manufacture

Like we discussed in small-molecule drug manufacture, there is also a focus on bringing continuous manufacturing principles to biologics. One example we briefly discussed was the modification of bioreactors to support continuous perfusion of nutrients to cell cultures in order to expand yield beyond the amount in the initial medium. These types of reactors are especially great for biologics that are secreted by the cells because the medium is continually harvested and replaced allowing for extended batch runs. These are ended after a predetermined time (potentially weeks or months later) in order to prevent cellular mutations that may affect drug quality. Automation around filtration and purification also makes the process of switching through different filters and chromatography types very efficient.

• Closed and Automated Systems

Closed systems close off the manufacturing process from the open air greatly reducing contamination and making cleaning easier and more efficient. Automated systems work well with this and can integrate multiple steps of the manufacturing process without human intervention.

References:

• Biopharmaceutical Manufacturing: Principles, Processes, and Practices: Gary Gilleski, Charles Rutter, Becky McCuen

• Novo Nordisk (Ozempic): Acquired Podcast

• History of Insulin: PMC3714061

Since the earliest human records on medicine in the Mesopotamian civilizations of Sumer and Babylon we have sought to manipulate the course and progression of human ailments with external substances found in nature. These civilizations and most since have lacked the comprehensive scientific understanding of biology, chemistry, and physics we have today and thus where progress was made, it was by happenstance. For example, in records we have from Egypt, willow bark was used to treat pain, inflammation and fever. In the 19th century, scientists were able to extract salicin from the substance and create acetylsalicylic acid, or aspirin. While impressive that ancient civilizations stumbled upon effective remedies, these were often far and few between. Taking centuries or millennia for new effective remedies to be found and to weed out the cruel, unusual, and ineffective remedies that were even more common. For example, just two centuries ago, it is believed that the practice of bloodletting was a significant contributor to George Washington’s death from a throat infection.

The evolution of the field of biology led by the invention of microscopes, discovery of cells, and the explanation of the role of DNA has led to an explosion of progress within the medical sciences. Scientific theories are explanations as to why certain phenomena arise in the world. They assert the existence of real things and processes that can be tested by the result of an experiment. Thus, they form a scaffold for future research and provide a mechanism for their own overthrow by future competing theories. When Darwin formulated his theory of evolution, implicit in his theory was some way for information to be passed down from generation to generation even though he acknowledged in his Origin of Species that the exact mechanism was unknown. However, armed with the knowledge that something must exist to fulfill this role in the theory, the theory of information storage in organisms was developed. With the unit of selection found in the genes encoded by DNA, we had for the first time a foundational understanding of how observed phenomena in the body arise.

Since the 20th century, with a better understanding of cells and biological processes, we’ve dreamed of powerful ways to counteract abnormal states in the body. Pharmaceuticals and biologics began to appear and accelerate rapidly. When this coincided with the information and computer age post-WWII, we dreamed of powerful microscopic machines that could live in our body and counteract damage and decay in real time. In popular culture and media, these appeared as scaled down machines, like little ships that could navigate the body.

In this post and two follow-ups I will provide a primer on the three major categories of biotechnology that I alluded to above: drugs, biologics, and nanotechnology. For each, I will outline what they are and how they are different from each other, how they have evolved over the years with a case study on a flagship example, and to touch on some of the modern innovations around their development. My goal is that these primers serve as a scaffold for future posts within the biotechnology domain.

Definitions:

Small Molecule Drugs:

• Chemically synthesized molecules that are designed to interact with specific biological pathways in the body.

• Some otherwise defined biologics like antibiotics are considered small molecule drugs because of their small size, relative simplicity to create, and major differences in their development processes from other biologics

• Examples include aspirin, antibiotics, and statins

Biologics:

• Substances that are derived from living organisms or their cells

• In some cases biologics can be whole cells like in the case of engineered T cells

• Other examples include hormones like insulin, antibodies, and vaccines

Nanotechnology:

• The creation of materials or substances like nanoparticles that can perform functions on the 1–100 nanometer level in cells — like delivering particles into cells or repairing damage. For reference a DNA double helix is around 2nm in diameter and a water molecule is .275 nm

• Examples include the nanoparticles that carry mRNA vaccines and other therapeutics into cells.

Small Molecule Drugs:

The concept of drug use, that is external chemical compounds being administered to the body, spans as far back as written history. Initially, these were based on rules of thumb that certain natural substances appeared to have a given effect. The next big breakthroughs consisted of isolation of the active ingredients within those natural substances. Beginning in the 1800’s, the scientific field of chemistry led to the first isolations and synthesization of these compounds. In 1805, the chemical morphine was able to be isolated from opium and by the end of the same century in 1897, aspirin was artificially synthesized by Bayer. This synthesization marked the beginning of modern pharmacology. Throughout the 20th century, advances in scientific understanding of biology and chemistry led to the mass production of antibiotics like penicillin, statins, chemotherapeutics, and many more.

The defining characteristic of drugs, also known as small-molecule drugs, is they are able to be isolated, synthesized, and modified. Like we’ll discuss with the biologics, the first methods of creating these compounds were highly inefficient and revolved around isolation of substances from naturally occurring processes.

The History of Aspirin

The naturally occurring precursor to aspirin, salicin, was originally found and extracted from willow bark. Salicin is eventually converted by the body to salicylic acid which blocks the production of prostaglandins and thereby inhibits pain, fever, and inflammation responses. Extracting salicin for pharmaceutical use involved a multi-step process of grinding up the bark, dissolving the compounds in water or ethanol, filtering out unwanted insoluble matter like cellulose, and several rounds of crystalizing/precipitating out the salicin from the rest of the mixture to purify it. In 1829, the French chemist Henri Leroux isolated 30 grams of purified salicin from 1.5 kilograms of willow bark.

This process, while a major milestone for pharmaceutical science, was far too inefficient for commercial use. Additionally, the pure salicin had side effects causing gastric distress which led to a desire to modify its structure. Chemical experiments were done in the 1850’s by another French chemist Charles Gerhardt to add an acetyl group to salicylic acid to form acetylsalicylic acid (ASA). His process, more of an experiment, did not produce a stable or pure substance and was never commercialized. At the same time, chemical synthesization of salicylic acid via the Kolbe synthesis using sodium phenolate finally liberated the process from natural extraction from willow bark. Building on this in 1897, German chemist Felix Hoffman at Bayer developed a scalable solution for creating ASA using acetic anhydride to add the acetyl group to synthesized salicylic acid, mass producing what became marketed ever since as Aspirin.

Small Molecule Drug Manufacture: A Case Study on Aspirin

Taking aspirin as an example, let’s look at the evolution of the manufacture of small molecule drugs. Since at the end of the day chemistry is chemistry, generally the overall requirements of the process in terms of reactants and reactions will remain the same. The primary differences we’ll look at are the technology that enables this process. In the historical Bayer process, most of it was manual. The reactants (salicylic acid and acetic anhydride) were added in measured quantities to large glass or iron vessels along with a catalyst of either sulfuric or phosphoric acid and heated to 90 degrees celsius. The ASA was then dissolved in water or ethanol and repeatedly manually cooled, crystalized, and filtered to get as pure as possible. To dry the crystals, they used either ovens or let it air-dry. Aspirin was originally sold as a powder but when it switched to a tablet form, this powder was mixed with binders like starch and pressed into a mold manually or through steam operated machines. Quality control was limited to visual inspection or testing of a small subset of batches. In the end, batches were quite small by today’s standards given the equipment they had and could take hours or days to complete given the highly manual processes involved.

Contrasting with this process, are our highly automated and technologically advanced drug factories today. Recently, the concept of continuous manufacturing is taking off to drastically increase yield and efficiency. To enable continuous manufacturing, the process has become highly automated. The automation helps to control for things like quantity, temperature, reaction speed, and other sources of human error. With aspirin for example, reactants might be continuously added into a machine called a plug flow reactor where as they progress along the tube, the products of the reaction are formed. The plug flow reactor allows for continuous production because it is constantly being fed reactants and the progression through the system creates a precise time interval for the reaction in order to guarantee high conversion. (Conversion is the measure of the reactants successfully coming into contact to form the product). The addition of reactants and catalyst is controlled and precisely metered by automated systems to avoid any human error in chemical concentrations.

After the initial reaction the crystals make their way to a number of additional machines that would make the early Bayer chemists envious. Instead of manual filtration and purification, large scale systems filter out the aspirin from the impure slurry produced previously, send it to be evenly cooled and crystalized, and then to a centrifuge to further separate the crystalized aspirin from the rest of the mixture. From there, the crystals are dried in trays under circulating air or by suspending the crystals in hot air within systems that continuously monitor heat and moisture content. In modern factories, these machines are typically connected to allow for continuous manufacture and can be automatically tested for quality control with integrated tools like inline spectrometers. These inline spectrometers shine light onto the sample, collect data on how it is being absorbed, refracted, or scattered and run the data through predefined models to automatically test and monitor for deviations in purity, concentration, or other metrics. Finally, automated tablet pressers make thousands of tablets per minute and coat them with a fine spray in a tumbling drum that will assist with swallowing the drug.

Conclusion

As we can see, a lot has changed in the history of drugs and drug manufacture. In stages we learned to:

1. Consume willow bark for it’s salicin content that helped with pain and inflammation

2. Extract the salicin from the willow bark

3. Chemically synthesize the salicylic acid that your body converts salicin into

4. Efficiently modify a molecule (salicylic acid) with an acetyl group to to ease side effects

5. Automate nearly the entire process to supply the world with cheap aspirin

We also saw in this case study the emergence of continuous manufacturing principles in small drug manufacture which emphasizes automation and continual flow through the process. This makes the process much more efficient and has a big impact on increasing yield. In a healthcare system dominated by high prescription drug prices it becomes more and more important to focus on increasing efficiency and supply to help drive costs down for everyone. In the next primer, I will turn my attention to biologics and dive into another case study to demonstrate key principles around their development.

Subscribe now

Introduction:

Science fiction is a genre known for pushing forward the boundaries of human creativity in technology and opening up new conceptions into what a future world could look like. Because of the desire for narrative conflict in our stories, science fiction, like all genres of storytelling, requires an antagonist. Science fiction I’ve seen generally finds part or the whole of this antagonist either within it’s fantastical technology, or it sets up the advanced technology as a contrast to a horrible external foe to the human race or specific social classes. In cases where the world for the average person is worse than today, we consider these worlds to fall into the class of dystopias.

Although humans crave a narrative antagonist, we also crave a resolution to the conflict of a story. In science fiction dystopias, this resolution is often achieved by either destroying the offending technology or finding a way to cope and live with the conflict from technology or external forces. Even though a resolution is obtained by the end of the story, the state of the human race is often worse than today’s real world, and the implicit message in many cases is that technology is harmful at worst or impotent at best in solving humanity's most pressing problems. Given the influence of popular culture on how people think about their everyday life, I believe these stories generate a sort of subconscious anti-human, anti-technology viewpoint that influences decisions up to the policy level. We can see it directly today with AI, where the potential dangers of AI are pulled right out of science fiction stories.

What I want to discuss in the rest of this piece is the exact opposite view. What I see when I watch these science fiction stories depicting dystopia are worlds that are so illogical that they border on impossibility. In almost all cases there exist extreme logical inconsistencies with the simultaneous existence of fantastical technology and relatively trivial problems that a world capable of these fantastical technologies could easily solve. In both science fiction and real life, people struggle to identify cross-applicability of knowledge and technology as well as how human incentives dictate the use of that technology. When one starts to look at the broader dystopian world with this in mind, we find the villain is nothing more than a paper tiger. 

Failure to understand how human knowledge builds into a unified whole is perhaps the biggest reason why these worlds seem so shockingly inconsistent. Since the scientific revolution and our first good physical theories like those of Newton, scientific progress has built up a scaffold of knowledge that crosses every boundary of human knowledge today. Without an understanding of electromagnetism, boolean algebra, and a host of other mathematical and physical theories we could never have created computers. Despite this, none of these were developed with computers in mind. As we accumulate knowledge of how the world works, we can apply this knowledge to build new and better technology in other domains. Since humans are fundamentally problem solvers, we will always search for tools to solve our pressing problems no matter where these tools come from. Therefore, the assertion of the existence of an advanced technology necessarily asserts capabilities of both that technology and the broader world based on all the scientific knowledge that enabled its creation. This will be a key point in understanding why many of these worlds are logically inconsistent. 

With the understanding that these stories are fictional and many contrivances are made so that the plot can occur, I am going to criticize several popular science fiction dystopias. We will look in each case at the broader state of the world, the nature of it’s dystopia, and the technology at humanity’s disposal. In doing so, I hope the reader comes away from future dystopian stories with a framework that reduces anxiety and allows the enjoyment of the art without taking for granted its assumptions. This process is how I learned to stop worrying and love sci-fi dystopias. 

Interstellar

The World and Plot:

The world of Interstellar is set some decades into the future in a world that has in many ways been reduced to the past by catastrophe. In this world, a progressive plant disease called “blight” is slowly killing off all crops and plant life that humans rely on for food. There is allusion to military conflicts over food scarcity but by the present day in the movie, humanity has all but abandoned any pursuit that is not tied to farming. The protagonist Cooper, a former NASA pilot and engineer is now a farmer, albeit a highly advanced one with a fleet of large programmed farming vehicles. As the years pass by, the blight evolves to infect more species of crops and the devastation it leaves on plant life enables massive dust storms that become a regular fact of life. The dust is shown to actively harm people, with some characters developing respiratory issues. And if all of this wasn't bad enough, it is revealed that due to the death of plant life, the Earth will run out of oxygen and destroy humanity.

I’m not going to go too deep into the rest of the plot, but the solution that the remnants of NASA come up with is to send out astronauts through a wormhole to an area of space with several possible planets to resettle humanity onto. Though I have my critiques on the realism of that section of the movie I want to tackle the premise of the dystopia the writers have made. Below I outline the main “antagonist” of the movie as well as the technology that we see humanity has developed thus far. After that section, I’ll attempt to show how even this small set of advanced technologies as well as what we are already capable of today (this movie is set decades into the future) invalidates much of the conflict that we see in the movie. 

The Dystopia:

• Blight - a plant disease that is slowly killing off all crops and plant life and reducing the Earth’s oxygen to uninhabitable levels. The lack of food and oxygen is the primary driver of the conflict of the story.

The Technology:

• Advanced, potentially general, artificial intelligence (ex: TARS, CASE)

• Advanced spaceships and space faring technology (life support, advanced materials, etc)

• Perfected cryo-sleep 

Why This Wouldn’t Happen:

A good place to start when breaking down the conflict of this movie is with its root cause, the blight that is destroying crops. This presumably fungal disease has created many downstream problems including wars, reduced population, and diminishing oxygen in the atmosphere. Fungal diseases affecting crops are not new in today’s world where we use a mix of things like fungicides, genetic modifications, and crop rotation to fight it. However, let's assume that this fungal infection is unusually aggressive and fast mutating. Even with this assumption, we see that it is not necessarily THAT quick evolving since it's been around for decades in the movie and it seems to take at least a few years before it evolves such that it can jump to a new crop type. 

What do humans have to defend against this? When thinking about a war between humans and biology, it's helpful to understand what we are fighting and what we are fighting with. A biological enemy like a fungus does not think, it is governed by evolution which has no direction in mind. Evolution is a process where random mutations occur and are trialed in an environment where the less advantageous genes cause their hosts to die or be outcompeted by rival genes. Though this sounds scary, it's actually an inherently slow process governed by random chance. Most mutations that happen have no effect one way or the other and a mutation that harms the host or the species is just as likely as one that benefits it. Contrast that with humanity’s own trial and error mechanism: knowledge creation. Though it also relies on testing ideas against reality to find which one works, it DOES have a direction in mind and is creative in a way evolution is not. This means that memetic evolution (evolution of ideas) is much much faster and more capable of solving a problem than genetic evolution. 

The first line of defense in this war would be biological countermeasures using today's technology and theirs. Interstellar was created in 2014 so perhaps it is unfair to them that they didn't quite know about the work being done to develop CRISPR (a highly advanced gene editing technology). At the time of that movie, the predominant gene editing methodology was done either by selective breeding or recombinant DNA, both of which are orders of magnitude less advanced than CRISPR. CRISPR is significantly faster and cheaper than those traditional methods, meaning it is a biological superweapon that the writers of Interstellar never counted on. 

This blindspot to technological development is a common theme and leads to pessimistic views of the future generally. It is much easier to imagine a problem than to creatively develop its solution. In this case, the story writers imagined a world where we were capable of perfected cryo-sleep, complete preservation of human cells including the brain, able to be frozen and thawed at will to withstand any aging for years during a journey through space. But a world that is capable of developing that technology is one that has so mastered biology that it is hard to comprehend. It requires understanding of starting and stopping life at will down to every metabolic and cellular pathway that could be effected. We are nowhere close to that level of biological mastery yet we already possess the technology required to defeat the enemy of blight – advanced gene editing. 

Advanced gene-editing harnesses the power advantage that humans have over a natural pathogen like blight. That advantage is the speed of directed memetic evolution versus natural evolution. We see in the movie that the blight has been around for decades and evolves relatively slowly to new targets. This is because it has to undergo an undirected process relying on random chance that a beneficial mutation will occur that allows it to jump to a new species. Since this takes years at the least, humans are at a natural advantage. Even with today’s technology, which is woefully behind that of the world of Interstellar, humans have shown remarkable speed in developing biotechnology under pressure – notably developing an effective vaccine to a novel virus within 1 year during the covid pandemic. Given that a mastery of cryo-sleep technology displays such advanced knowledge of cellular biology, it is very plausible that humanity in this scenario would be able to very quickly out-engineer the blight. This could be: genetically modifying the crops, genetically modifying fungal predators (other fungus, viruses, bacteria), genetically modifying the blight itself, developing targeted fungicide or plant vaccines, or all of the above. 

It's even worse than that however. Even taking as a premise that humanity cannot defeat the blight pathogen via genetic engineering, it possesses other extremely powerful tools in its tool belt. Namely, it is shown to possess extremely advanced science in: life support, materials, and artificial intelligence. These combine to make the case that humans can either create massive greenhouses that are fully automated, airgapped, and run by AGI they’ve already demonstrated they possess OR they can build massive interconnected life support domes for themselves to live within. This civilization is already capable of advanced space travel and the challenges of constructing these facilities to live in shouldn't be harder than the spaceships they've already demonstrated being able to build. I would be remiss in mentioning that the level of artificial intelligence they’ve demonstrably created in the movie would make it even more likely they could out-engineer the virus. Today’s far less advanced AI was vital for developing the covid vaccines as quickly as they were. 

Lastly, the storywriters require a suspension of belief in the very principles of biology itself. Nature is, everyday, in a biological arms race between competitors, predators, and prey. A disease like blight would create extremely intense evolutionary pressure for all plant life domesticated and wild. It is very likely that a natural equilibrium would eventually occur especially amongst wild plant species which are much more genetically diverse than domesticated crops. This natural equilibrium within the wild plant species would likely solve the pressing problem of diminishing oxygen levels by itself. Additionally, the widespread prevalence of a fungus that is outcompeting existing fungi for resources would put evolutionary pressure on fungal species to potentially attack the blight causing fungus itself given that fungi are the primary predators of other fungi in nature.

WALL-E

The World and Plot:

The movie Wall-E made by Pixar is perhaps one of the most popular movies among my generation when alluding to potential dystopian futures. In the world of WALL-E, a mega corporation called Buy’N’Large came to dominate the world on Earth. Due to overproduction and consumption, the Earth becomes literally buried in garbage, all plant life goes extinct, and toxic air makes the earth uninhabitable by humans. The solution that the company comes up with is to move humans temporarily to luxurious space stations in orbit and leave the earth to be cleaned up by robots in the meantime. When this cleanup effort fails they make the decision to never return to Earth.

The movie is set 700 years after this decision and humans on the main space station of the story, the Axiom, have essentially been reduced to obese consumption machines floating around on anti-gravity hover chairs. They show no interest in anything that isn't mindless consumption and are kept alive by an extremely advanced city scale space station run by an equally extremely advanced artificial intelligence and robotics network. These robots handle everything to do with maintaining the vast space station and its inhabitants. The story follows two robots WALL-E and EVE as they fight the artificial intelligence that runs the ship and the apathy of its human inhabitants to get them to return to Earth. 

As before, I am less interested in the plot as it proceeds than I am in the backstory that created the dystopia in the first place. 

The Dystopia:

• The Earth is over run with garbage and pollution forcing humans to abandon it and live in space stations

• Humans are too apathetic and concerned with consumption to care about the environment

The Technology:

• Artificial general intelligence (ex: EVE, WALL-E, and AUTO)

• Artificial gravity

• Autonomous manufacturing

• Advanced energy systems – likely nuclear fusion and self-maintaining power systems

• Advanced robotics, computing, and sensors

• Advanced life support systems

Why this Wouldn’t Happen:

Once again, as is almost always the case with these stories, the writers have created a world of contradictions. We see a human race capable of unfathomably advanced technology by our standards get reduced to consumption slaves aimlessly circling an earth that has been destroyed environmentally. To understand exactly how implausible this outcome is, you have to fully understand the premises of the film. In the film, humans have amassed pollution that buries entire cities. They also imply that humans had already abandoned pursuits that weren't consumption based as that is the only way to explain how the world gets destroyed. However, it’s required that at the same time, humanity would have needed to be developing the extremely advanced technologies that they later rely on to escape Earth and survive in space. As we’ll see, that very technology would have saved Earth far before it needed to be evacuated. 

Perhaps the single biggest unexplained premise is the Axiom’s power systems. To sustain all of the technology on board between the computing, robots, life support systems, artificial gravity, advanced propulsion, autonomous manufacturing, etc for 700 years, it by necessity needs power systems that far exceed what we can do today. Given the energy needs and time duration we see it is very likely the ship relies on nuclear fusion technology or energy sources like solar or fission that are autonomous and self-reparing. Energy is the bottleneck for doing just about anything advanced in the universe, so humanity possessing this level of energy generation changes everything. 

Even today, the primary bottleneck for the feasibility of fighting effects of climate change like carbon removal or desalination is energy. A world with abundant and cheap energy can do anything that isn’t outlawed by the laws of physics. In the world of WALL-E, it implies that humans for whatever reason, at the same time they were developing extremely advanced energy systems, decided not to deploy that to help the environment around them. Even stopping here we know how implausible this is. It’s an economic fact that as countries get richer they decide to place more and more importance on the environment and the pollution they are generating. This relationship is modeled by the Environmental Kuznets Curve where developing countries grow economically alongside their pollution until pollution peaks at middle income. From there pollution levels decline. 

The hypothesized reason for this is that low income countries are primarily concerned with immediate needs and do not have the wealth or luxury to focus on secondary needs like the environment. As a society gets richer and the basic needs are met, the population shifts its attention to issues secondary to immediate survival, and in this case, environmental safety. Since they are also wealthy societies, they can afford the technology and work required to do so. In the United States, as we grew wealthier post WWII, the population demanded the passage of environmental protection legislation like the Clean Air and Water Acts. As a result, today, the United States has comparable or better population weighted air quality than most of Europe and the OECD (proxy for developed, high income countries). We would hardly call the United States less capitalistic or consumerist than its European counterparts, in fact we would probably say that the United States stands out from the rest in this regard. The premise of the story that the environment gets worse as societies get richer completely and totally fails to be supported by any empirical data. 

As a reader you probably can see the writing on the wall for the plausibility of the premises at this point. An extremely wealthy society that is essentially post-abundance, with limitless food and energy, and has built the capacity to manufacture a fleet of autonomous intelligent robots, has apparently stopped caring about whether trash is on the street or if their air is breathable. Everything we know about humans and economics points against this outcome. And everything we know about their technology points to these problems being easily solved. If they do have advanced fusion energy they already have limitless clean energy that is producing no pollution and can easily perform feats around carbon removal, desalination, and pollution clean up. If they have a fleet of extremely advanced robots and recycling capabilities (they were self-sustainably maintaining their space ship for 700 years) they could easily get rid of or repurpose the physical waste. 

Like most dystopias, this one again points to a narrative the writers wanted to portray rather than a real possibility. 

Fallout

The World and Plot:

The world of Fallout as described in the video game series and recent TV series by the same name is one which diverged from our current timeline around 1950. In this world, nuclear fission technology became widespread and transistors were never invented leading to bulky electronics relying on vacuum tubes. This creates a retro-atomic age aesthetic that seems almost stuck in what we see as the 1950’s of our timeline, except highly advanced. 

In the Fallout world, the Earth’s energy supplies of oil and uranium powering their societies are rapidly dwindling and creating global conflict. This caused a world wide global war from 2052-2077 including wars between Europe and the Middle East, a Chinese invasion of Alaska, the annexation of Canada by the US, and a subsequent counter invasion of China. This culminated in what is called the “Great War” in 2077 where the US and China destroyed the world in a nuclear exchange. 

In the time after diverging from our timeline in 1950, the world of Fallout rapidly advances in many technologies beyond what we are capable today. Essentially everything becomes powered by nuclear energy (likely fission first and then fusion) from appliances, cars, robots, and whole buildings/cities. In the official Fallout lore, fusion technology is not technically invented until 2066, or 11 years before the nuclear war. However, we see that fusion became remarkably widespread in that time, where pretty much everything is powered by it in day to day life by the time of the Great War. 

In a demonstration of engineering prowess, the company Vault-Tec builds over 100 vaults across the US for people to live in in case of nuclear war. Most of the story lines explored in the Fallout Universe consist of someone from these vaults interacting with the irradiated wasteland, its inhabitants, and various warring factions vying for power. The US 200 years after the war is shown to be an inhospitable wasteland with mutated humans (termed ghouls), mutated animals, and undrinkable water. In the vacuum of power, powerful factions are constantly warring, nobody can keep the peace against bands of mercenaries and gangs, and the regular folk suffer. 

The Dystopia:

• The countries of the world are locked in furious competition for energy resources that leads to conventional and later devastating nuclear war

• The resulting nuclear war has reduced humanity to a semi-primative state and the environment around them into a harsh wasteland for at least 200 years as shown in the various installments of the franchise

The Technology:

• Nuclear fusion- including small scale fusion cores

  • Powering cars, appliances, and entire cities and buildings 

• Advanced autonomous robots

• Artificial General Intelligence - ZAX supercomputers

• Perfected Cryo Preservation

• Incredibly advanced bio-nanotechnology (instant healing stimpaks)

• Personal invisibility field generators

Why This Wouldn’t Happen:

This dystopia, while relying on cartoonishly evil companies that think nuclear war would be more profitable for them (absurd on its face), is mostly about resource constraints and is a common theme among dystopias. I think it's helpful though to understand that the context around the development of the Fallout franchise in the late 1990s consisted of a time of dramatic resource pessimism. Common predictions placed peak oil in the early 2000’s and uranium depletion by 2013-2015. Today’s estimates place our known conventional uranium reserves as capable of sustaining 100 more years of current demand with even more available in unconventional deposits and seawater. Additionally, Thorium, 3x as abundant as uranium, is a known replacement for uranium to supply nuclear fission and the Chinese recently developed a thorium reactor. But this still understates nuclear fission’s resource availability. Breeder reactor technology that has been around for many decades is capable of recycling spent fuel and is estimated to be capable of extending our supplies for another 30,000 years. Oil on the other hand has known reserves of 47 more years but has continually increased as we’ve developed new technology to access more than before. Hydraulic fracturing technology and horizontal drilling for example has drastically increased the supply of natural gas and shale oil previously thought inaccessible.  

Why are resource pessimists consistently proven so wrong? The constant theme is that they underestimate or do not even try to estimate the growth of human knowledge and the effects of economics on human incentives. A resource is nothing more than a material that can be transformed into something else with the right knowledge. To paraphrase Naval Ravikant: “To a caveman, very little can be considered a resource”. For tens of thousands of years, uranium has sat inert in the Earth around us and was not considered a resource at all. It only became useful when we invented technology that could turn it into a resource. Shale oil and natural gas was not a resource until we invented technology to extract and use it. Assuming a world where humans do not invent solutions to their problems will always be fatal for predictions of the future. Thomas Malthus famously demonstrated this by predicting population collapse due to lack of farmland – completely missing the innovations that have allowed us to use even LESS farmland and feed over 7 billion more people than when Malthus warned of the population limits of humanity. 

From an economics perspective, the incentives for different energy sources and new technology change as prices change. If uranium becomes much more expensive due to dwindling supply, it becomes more viable to invest in the technology required to extract it from seawater, switch to thorium, or use breeder reactors to recycle the fuel. As oil becomes more expensive, it becomes more viable to invest in finding new deposits, developing new extraction technology, or switch to other cheaper energy sources like solar. As other energy sources like solar, wind, nuclear, geothermal, etc come online, the demand for oil drops and the amount of time the reserves can be used is extended. When it comes to energy though, the only timeline relevant with regard to resources is the time it takes to develop nuclear fusion. 

To the credit of the Fallout creators, they didn't completely ignore the growth of human knowledge. Miniaturized nuclear fusion was invented 11 years before the nuclear war. Nuclear fusion, the process by which stars generate energy, effectively provides infinite energy – billions of years worth just using known resources on Earth. Fallout’s timeline is pretty unrealistic though since in those 11 years miniaturized fusion became ubiquitous in society, powering pretty much everything. In real life, fusion would have been useful as large scale reactors far before we could develop a hand held fusion reactor like in Fallout. Today we have already demonstrated remarkable progress in nuclear fusion, achieving net energy gain for the first time in 2022 at the Lawrence Livermore National Laboratory in California. The promise of nuclear fusion has never been more real and we are nowhere near depleting the resources needed to sustain our current and projected lifestyles – especially as our switch to new technologies like solar reduce demand for resources like oil. 

To wrap up, the Fallout universe completely ignores the reality of energy technology advances while at the same time proposing incredible advances in other fields. They develop instant healing stimpaks that can instantly heal all wounds, presumably through highly advanced nanotechnology. They develop miniaturized nuclear fusion/fission demonstrating extremely advanced engineering and physics prowess. They develop highly advanced robotics and artificial intelligence, perfected cryo preservation, invisibility field generators and more. All far more technically challenging than nuclear fusion or much simpler technologies like breeder reactors, uranium extraction from seawater, or thorium reactors. If I can imagine ways in which a society with those technologies can solve its problems with their own abilities, it is literally unbelievable that they couldn't when faced with nuclear war over the issue. Knowledge compounds and applies across fields. It is virtually impossible to make extremely advanced technology in one field without creating applicable knowledge to another. 

Conclusion:

This piece was written to inject some much needed optimism into the culture around science fiction and technology broadly. It has become far too popular to be pessimistic about technology and humanity’s ability to solve our problems. In my opinion much of this can be blamed on the media, both popular and fictional, in emphasizing a pessimistic view of the future, technology, or human abilities. For my entire life I have been surrounded by a deeply ingrained view that the world is worse than it was and will only get worse still. For these people, science fiction dystopias allow them to visualize how this view is the correct one and that humans really are helplessly and inevitably creating their own destruction. Instead, the optimists have been quietly making the world better and better through nothing other than incentives and desires to solve problems. Yes, problems are inevitable, but they are also soluble. It just requires someone out there dedicated to creating the knowledge to solve them. 

References:

• https://www.greenpeace.org/static/planet4-india-stateless/2024/03/44a856c8-2023_world_air_quality_report.pdf?utm_source=chatgpt.com

• https://www.llnl.gov/news/ignition

• https://online.ucpress.edu/elementa/article/doi/10.1525/elementa.405/112758/Democratizing-CRISPR-Stories-practices-and?utm_source=chatgpt

Introduction

The story of the United States of America is the story of the creation of the wealthiest country that has ever existed. In just a few lifetimes, since its start as the outcropping of British colonies, this country has increased 400% in size, 13,258% in population, and grown to absolutely dominate the world’s economic and military might. This rapid growth and prosperity has created an ethos around what is termed as the “American Dream”. The idea that anybody from any background can have their own slice of economic prosperity here. For a long time, this was taken for granted, and to many extents American citizens are still better off than their forefathers. For many however, something big feels missing. 

To understand this feeling, it's important to look back to see what has and has not changed in the past decades and what, if anything, has changed recently. Some of the most infamous charts in use today in the corner of the internet I frequent are below. They demonstrate that while certain products have gotten cheaper, many goods/services that Americans consider foundational to the “American Dream” have ballooned out of control. Alongside that, we see that manufacturing jobs have greatly declined (though output has increased or remained consistent), and GDP growth has slowed. 

Price Shocks

In the last three decades there has been a wild divergence in prices between goods and services for Americans. Consumer goods and leisure have stayed the same or decreased while goods and services tied to basic needs have exploded1. 

As we can see from the next graph, nominal wage growth in the same period (130%) has kept up with some categories’ nominal price increase. For tuition, day care, and medical care, it has not kept up. For energy, housing, food, etc it appears to have kept up but it's important to note that the housing category used in the above graph includes utilities, household operations, and furnishings which grew much slower than rent and home price. However, according to the national home price index, the cost of purchasing a home has far outrun wage growth at +270% over the same period2. Rent has outpaced wage growth at +~164% over the same period3.

Aside from whether or not wages have kept up with inflation of price, something feels deeply off about prices increasing at all. Americans have come to expect that competition and innovation will lower prices the way that has worked out for electronics. The TV in my living room today would have cost easily 3-5x as much when I was a kid. To some extent, Americans are not wrong to think this way. If innovation can give us an affordable phone with more computing power than the computers that sent astronauts to the moon, why can’t it scale this way for other industries?

Supply Shocks

Wage growth and price inflation is only one part of the story however. If you are like me, the Covid-19 pandemic was a massive wake up call for the preparedness of our country in terms of supply chains and critical goods. As an early Gen-Z I grew up in a world that was already well underway with the globalization that began to occur in the 1980s and 1990s. For people like me and even older I think we took for granted the supply chain systems that enabled the global system and were shocked that we struggled to keep our hospitals supplied with basic necessities like masks, PPE, and ventilators.

Since then, the topic of manufacturing has entered the American zeitgeist with force, where many people point to the below graph of the decline in the percent of American workers employed in manufacturing. Their aim is to show how globalization has stripped many aspects of the industrial supply chain out of the United States as factories shifted to other countries like China in order to save money with labor arbitrage. 

Critics of the burgeoning re-industrialization movement in America however, like to point to the below graph indicating that while employment in the industry fell, industrial output has remained roughly the same and appears to be in a potential upswing. In their view, automation is the primary reason for the loss of jobs and not globalization. However, this view, in my opinion, makes a few mistakes.

First, industrial output is measured in the value of the goods being created. So while output appears flat, it is resting on a few industries that produce very expensive goods like chemicals, transportation, and advanced electronics which account for roughly ~40% of our total manufacturing output. These critics in turn respond that this is a feature rather than a bug and that we should be focusing on “high-value” production. But this leads to my second point. 

Manufacturing output (monetary) as your primary metric greatly underestimates the strategic importance of industry. As we saw during the Covid-19 pandemic, we rely on other nations, and primarily an adversarial one in China, for many key inputs and finished products that are critical for the health, well being, and strategic self-sufficiency of Americans. Mask and PPE (personal protective equipment) manufacturing would be very low value in the monetary sense that goes into the above graph, but is far from low value in a strategic sense.

Focusing on monetary value misses the forest for the trees when it comes to maintaining a world superpower. In traditional economics, the goal is to shift from materials and upstream precursors of manufacturing to finished goods, especially high value ones. This is because it is more “productive” work with higher margins than commodities like input materials.

But again, this has made the US reliant on a foreign adversary for many of our key upstream dependencies on high value goods as defined both monetarily and strategically. The vast majority of our chemical inputs to drug manufacturing, for example, come from China and other foreign countries. This makes the health of Americans directly dependent in the worst case on deliberate sabotage from an adversary and in the best case the whims of international supply chains that can fall apart at the drop of a hat.

In the third case, the implication of the critique that US manufacturing has lost jobs due to automation, is that US manufacturing must be very advanced compared to China. However, many manufacturing experts in recent years have been pointing out that China is rapidly overtaking us in advanced and automated manufacturing. China currently has 41% of all industrial robots in the world and accounts for 51% of all new installations4. This has made the prospect of re-industrialization even more challenging. Due to decades of building a manufacturing base, China has developed a breadth of expertise that has directly led to dominance across manufacturing as an industry. As US manufacturing expert Greg Koenig points out: “Having a progressive up-skill ladder in industry… is what accelerates the sexy, glamorous, high-value stuff at the top”.

Having an industrial ecosystem accelerates innovation and build out across entire industries. This is what made and continues to make Silicon Valley dominant in tech (the name comes from the chip industry that started there that was driven by the movement of talent among companies), and has confounded attempts at a nuclear energy renaissance. The United States recently finished its first reactors since the 1970’s at the Vogtle Plant in Georgia. The lack of expertise in building reactors due to this time spent not developing the workforce is cited as a large challenge by that industry. 

Last, critics of re-industrialization and attempts at de-globalization point out that globalization has led to GDP growth and in essence: “if it ain’t broke, don't fix it”. Yet we see in the below chart that GDP growth has been steadily slowing since globalization started. Between 1950-1974 the average growth was 4.06% per year, 3.28% per year from 1975-1999, and cut nearly in half to 2.22% per year in the 21st century.

To put that into perspective, the difference between the two growth rates is doubling the economy every 17 years versus every 32 years. If we had continued our trajectory of 4.06% since 1950, our economy would be the size of the US today plus the entire European Union. So while GDP is growing, it is growing at nearly half the rate than it was when the manufacturing workforce was 20%+ of the total workforce (~4 times higher than today). If we were truly moving people to more productive industries that whole time, why is our GDP growth slowing?

What Can We Do: Supply Side Acceleration

In this blog I will be focusing on supply side acceleration as a lens through which to make the United States a more dynamic industrial power and thereby maintain its position as the undisputed number one on the world stage. The name supply side acceleration was chosen to be a synthesis point between the libertarian leaning e/acc (effective accelerationism) movement and the “supply side progressivism” of the “abundance agenda” left. Supply side acceleration is focused on one thing: acceleration in the supply of our most critical goods and services. The most critical goods and services are those that are most critical for national security and for citizens’ quality of life.

While e/acc and abundance seem to share similar goals, e/acc has a bias towards government non-intervention while abundance, though framed to strategically focus government efforts, has ideological commitments added on. As Ezra Klein, the co-author of the book Abundance, said: “it’s not omnidirectional moreness. … I want more clean energy, not more dirty energy. If you come to me with a proposal for how we can build a shit-ton more fossil fuel plants, I am not excited about that. I will oppose you.” Supply side acceleration takes a pragmatic approach. If building more “dirty energy” in the short term is what is necessary to secure a better life for Americans and for American dominance in the fullness of time, I support it. Especially given the fact that what is ostensibly clean energy on a relative basis in natural gas, nuclear, geothermal, and hydroelectric gets regularly opposed on environmental grounds by the American left. 

Accelerating supply means bringing back critical supply chains and using technology, industrial policy, and strategic government non-intervention to unlock once again the industrial might of America. The goal is to re-awaken the American dream. Medicine, housing, and education will be cheap and abundant. High paying and meaningful jobs will be more available for the broad working and middle class. We can be the epicenter of big projects and new technologies that are unlocked by our industrial ecosystem and cheap power. You can’t predict the growth of human knowledge, what new technologies are created and where they can be usefully applied, but you can position yourself to best capture its benefits when it does grow.

The acceleration of advanced artificial intelligence technology is pushing our current system to a breaking point of its own making. The energy requirements for AI is on track to put a major strain on our energy sector made brittle by overregulation and supply chain deficiencies in the parts needed to build new powerplants and transmission lines. The output of AI, even if wonderful, is on track to be massively constrained by our ability to realize it. AI can produce the recipes for new drugs and materials, new designs for massive capital projects, or unlock the potential of robotics to eliminate the need for toilsome or unproductive work. None of this matters if we don't have the physical ability to create and scale those drugs, materials, construction, and robotics. To position ourselves for dominance in AI requires more than software dominance, it requires dominance in the world of atoms – the world of supply chains and industry. It requires us to build in the real world again. It requires supply side acceleration. 

Share

Footnotes