Antibiotic Alternatives
I have a new article out in the Earth Day 2024 edition of Utopia Science Fiction Magazine. Here’s a teaser.
We hear a lot about prolific organisms evolving resistance to our chemical weapons. Weeds and insects become resistant to our pesticides. Microbes resist our drugs. But how do they do it? When I teach evolution, students tend to confuse microbes with the Marines, who have an unofficial motto:
“Improvise, Adapt and Overcome.”
Marines learn, individually and in groups, by adding and subtracting ideas from the pool of information in their behavioral repertoires. The organisms I'm talking about learn as populations, by discarding individuals from their populations. Those who can't do, die. Learning happens, but the mechanism of learning is substantially different at different levels of organization, and those details are important for understanding how to interfere with the process of evolution for our own benefit.
To boost Utopia, whom I always enjoy working with, I include a link to The Sample, an AI newsletter recommendation service.
Rapid Response Cohort
Last week I mentioned a science policy fellowship program here in NC. It is specifically designed for people early in their careers.
Not long after I got back from my Fulbright Specialist trip to India, I applied to AAAS, the publishers of Science magazine and the largest professional society for science in the US (and the world?), to become one of their Science & Technology Policy Fellows, who use their scientific expertise to advise members of the federal government. It was a long and involved process. I made it to the semi-finals for the Judicial Branch, which only hires 1 or 2 people per year, but they didn’t like my policy brief, which suggested a way for judges to evaluate changes in voting laws for how they might increase or decrease the number of people who vote. I borrowed the idea from online retailers.
So I did not go to Washington to work for the Judicial Branch (which includes not just the Supreme Court but all federal judges — really, one person for all that?). And until this year, I ignored the many automated e-mail requests from AAAS to apply again.
This year there must have been some perceived urgency around Artificial Intelligence, and they opened up a special application, with a short deadline (which has passed). Below is exactly what I put in, minus my CV and contact information for references, and plus two links. As with the policy brief, I’m posting it so people have some independent idea of what does and doesn’t work, beyond the general Q&As that the program staff hosts with alumni. Call it a case study. The footnotes are additions for my paid Substack audience, where I have space beyond AAAS’s 7k character limit.
Reasons for applying for the AAAS STPF AI Fellowship Cohort.
I have long-standing interests in policy at an abstract academic level. This led me to apply for a AAAS Judicial Fellowship a few years ago (2019). I was a Semi-Finalist but did not ultimately get that position. This is my second try.
I have been following freshman Representative from North Carolina Jeff Jackson’s weekly updates on his adjustments to and frustrations with Congress1. He draws a distinction between behavior “for the cameras” and behind-the-scenes committee behavior, and consistently describes the Science and Technology Committee as his favorite part of the job. While he has been gerrymandered out of office and I would not have the opportunity to work with him personally, his take on government service has emboldened me to think that perhaps I can contribute.
Qualifications: A summary of your background and expertise – what you can bring to the fellowship.
I was broadly educated in molecular biology as an undergraduate and in many flavors of neuroscience as a graduate student. My interdisciplinary program at the University of Rochester was headquartered in the medical school at Strong Hospital, where there was an excellent clinical Grand Rounds program, open to PhD students as well as medical students. However, I also took courses and attended research seminars through both the Center for Visual Sciences and the Department of Brain & Cognitive Sciences, where people like Dana Ballard and Raj Rao were studying computer vision.
My PhD research consisted of building models of the human visual system from an engineering perspective, despite not having an engineering degree. As a post-doc, I went on to adapt2 methods for optimizing biological neural network models based on differential equations3, despite never having taken a DE course, and getting a C in Calculus for Life Science Majors as an undergraduate. Some might call this arrogance. I call it being willing to fail constructively. I have very little ego about facts, and I am unafraid to cold-call people and ask for their opinions or advice. Given that Congress consists almost entirely of lawyers with no professional-level scientific background, who are engaging questions they were not educated for, I feel like I might have an advantage in communicating with them.
I have continued this open-minded approach on in my careers as an educator and as a science & science fiction writer. I consider myself a synthesist, someone who recognizes patterns of behavior across the different fields of knowledge so thoroughly fenced in by our academic disciplines. At a large HBCU in NC, I worked with philosophers and engineers to develop a general education critical thinking course called Analytical Reasoning. Now I teach systems thinking and qualitative mathematical modeling to college freshmen. I am practiced at responsibly simplifying complex situations on the fly.
I am not myself an expert (much less a researcher) in artificial intelligence, but I do have a broad personal network of connections, including academics and entrepreneurs within these spaces. I can draw upon these people as resources.
Areas of interest: Outline the policy areas or issues that are of particular interest to you.
As the coordinator of my department’s internship program, I am especially interested in the effects of AI on employment and humane economic development. I have been reading Glenn Weyl & Audrey Tang’s book at plurality.net,4 where they reverse AI (‘artificial intelligence’) to IA (‘intelligence augmentation’). If the general goal of technological progress is to increase the productivity of labor, not to increase the accumulation of capital (by decreasing its expenditure on “excess” labor), then many of the employment doomsday scenarios evaporate.
Having recently read Annie Jacobsen’s NUCLEAR WAR: A SCENARIO5, I am reminded of the problems of autonomous systems in general. The entire point is often to make them react as fast as possible, which from a systems stability point of view is exactly the wrong direction. As shown by Donella Meadows6 for economic systems like supply chains, shortening response times leads to over-corrections, which further destabilize the system. As defense contractors push for automation of kill-capable systems at multiple scales (a push the Defense Department has so far resisted7) we need to think about the goals of the system. As with the employment example above, if we choose the wrong goal in designing a system, we can’t blame the system for doing what we designed it to do.
Role as a fellow: Describe how you imagine you might apply your scientific expertise in the legislative fellowship.
I am a neuroscientist by training, which means I have formally studied biology, chemistry, physics (especially complexity), and computer science, along with interesting bits of anthropology, linguistics, and sociology. Humans are interesting animals, with all the evolutionary baggage we share with all life on Earth. We add other layers of complexity through our large brains and our extended social processes, like culture and technology. Each siloed area of science has its own culture, its own technology, its own rituals. A young and interdisciplinary field like neuroscience is still composed of intellectual immigrants who are not completely acculturated, who recognize that the borders are artificial. This is a valuable perspective. I feel like I could be a bridge between specialists, an honest broker of information who doesn’t privilege particular viewpoints because of cultural preferences. Or at least does so less often.
As AI systems become more complex, they become more lifelike. We have hundreds of years of thinking about animal behavior, and decades of thinking about ecological systems to draw upon for analogies to the development of AI, if we choose to do so. So far most of our analogies have drawn on religion, as we imagine various versions of apocalypse, from Harlan Ellison’s SkyNet deliberately triggering a nuclear war in an attempt at self-preservation to Star Trek’s Vger getting its programming scrambled through “mating” with an alien space probe. These are emotion-based what-if fantasies, dressed up in technical language. More recently there have been lower-stakes explorations like Ted Chiang’s “The Life Cycle of Software Objects,”8 which could be useful.
My focus on science fiction examples is not a personal quirk. Precisely because most of our leaders are not academic scholars, who have not formally studied the underlying theoretical issues, pop culture has more of an influence over them than we might like to admit. Like them, I have been thrown across the surface of the pond, skipping from area to area. Unlike them, as a college lecturer and science fiction writer I have had the leisure to dig below the surface. I also have a large network of specialists at various institutions who I can draw upon.
I’m also much more used to writing for the public than most academics are. My CV includes two links to pieces on AI.910
Career goals: State your goals for after the fellowship and into the future, and how you envision the fellowship will help you achieve those goals.
I would like to continue my development as a writer and public intellectual. I have never worked in government, and there are whole sets of experiences I’d be able to write about first-hand. North Carolina now has its own much smaller STEM Policy Fellowship program, and while I’m too old to serve in that capacity, I would like to contribute at the state level, either here or in my home state of Kentucky. I might be able to replicate NC’s program there.
On a less personal level, I would share this experience with my students at the university level. Undergraduate education at this point is almost entirely theoretical. I am currently in the process of reviving my department’s internship program, which provides valuable real-world experience. Right now our partners are mostly local nonprofits and small businesses. Adding government and policy partners to that portfolio would be extremely valuable. A course in science policy at the undergraduate or graduate level is also an exciting possibility.
Speaking of Footnotes
One of my DMs offered me a copy of this book, which he found to be a real downer.
I can say from personal experience that it’s difficult to encourage people of enthusiasm and high imagination like entrepreneurs and SF authors towards accuracy and responsibility without sounding like a killjoy. I tried really hard in my PlotBot column for IGMS, using the same playbook of humor and humility as the Weinersmiths, and I don’t know how often I succeeded.
Anyway, one of the other players in that D&D group has been listening to the book on audio. It makes heavy use of footnotes, the whole point of which is to pull interesting but peripheral anecdotes to the periphery, where they don’t disrupt the flow of the author’s thought in the current paragraph. I don’t listen to much audio nonfiction, so I don’t know how most handle it, but having a single narrator break into reading me a footnote in the middle of a sentence sounds confusing. It might be OK if there were multiple narrators — one for the text, one for the references, and a third for the footnotes. That could work.
I’m only a little way into the book, but my spoilers who finished it already tell me that no, they really don’t address space elevators — or robots, which to me are the essential precursors to humans living off-planet. I didn’t think to ask them about genetic engineering.
While we were kibbitzing around the table about the futility of predicting the future in the face of technological development, our Dwarven bureaucrat’s player pulled a drawer out of the end of the gaming table. “I’ve been sitting here for years and never knew this was here!” We laughed, but this happens in science all the time. We don’t know what we don’t know.
Black Swan Biology
of Do Not Panic posted about just such a discovery a couple days ago.
They found clusters of neurons in the brainstem that appear to control inflammation, which is involved in possibly every disease ever. This may be a game-changer, and it was completely unexpected.
I wish I’d thought to mention it to my nerds. I would have looked so smart. But I was laughing too hard about the drawer thing.
Keep reading with a 7-day free trial
Subscribe to Doctor Eclectic to keep reading this post and get 7 days of free access to the full post archives.