Best wishes for the rest of 2023
Whoever and wherever you are — thank you for subscribing to “Securities” and reading us week after week. Please have a joyous and restful holiday season, since as I wrote last week in Extreme Epoch, 2024 is set to be a watershed.
We’ll be on hiatus on Dec. 30th.
Best of the Podcast 2023
2023 was an incredibly busy year, and nowhere was there more fervent attention than on artificial intelligence. OpenAI launched ChatGPT at the very end of 2022, and its implications found purchase this year among more than one hundred million users and the regulators who serve them. Billions of dollars of venture capital flowed into the AI space, with investors funding everything from data infrastructure and better model training to the applications that are already beginning to transform industries across the world.
Our final “Securities” episode this year is a narration of our 9 favorite shows on AI, which we titled 🔊 “WTF Happened in AI in 2023?” It’s the perfect way to end the year — thanks to our producer Chris Gates for compiling all this material together. The 9 episodes we included material from are:
- “That’s 100% what keeps me up at night”: Gary Marcus on AI and ChatGPT — Gary Marcus is an incisive critic of modern AI technologies, and this episode still forms the intellectual foundation of everything that is wrong with foundation AI models.
- May the AI be ever in your favor — our own Grace Isford attended the first AI Film Festival hosted by Lux’s RunwayML and talked about the state of the art frontiers of generative film.
- Chatphishing, veracity and “two years of chaos and a reset” — Josh Wolfe and I discuss AI and security, specifically the tactic that Josh dubbed “chatphishing” to describe using AI bots to defraud an individual by pretending to be their loved ones or perhaps a famous individual.
- First impressions of OpenAI’s new GPT-4 AI model — Grace and I discuss the big launch of OpenAI’s GPT-4 model as we experiment with this new technology.
- “Smell can be art, and it also can be science”: AI/ML and digital olfaction — Smell is just important to AI’s future as text, audio and video. That’s where Lux company Osmo comes in. Founded by Alex Wiltschko, Osmo is building AI models of scent that can help transform industries as disparate as flavor and fragrance as well as mosquito repellant.
- Lt. Gen. Jack Shanahan on rebuilding trust between Silicon Valley and the Pentagon — Jack Shanahan was the founding director of the Pentagon’s Joint Artificial Intelligence Center, where he handled the eruption of protests around Project Maven at Google. The key lesson learned? Transparency — from the corporate side.
- AI: Disruption, Regulation, and the Road Ahead — A companion to “Reckless Regulators,” I sit down with TechMeme Ride Home podcast host Brian McCullough and Supervised editor Matthew Lynley to talk about how the U.S., U.K., and E.U. are all coming together to squelch AI innovation.
- Techno-Pragmatism: Looking Beyond Blind Optimism and Hopeless Pessimism — Josh and I talk about how to balance aggressive optimism about the future of technology with its pernicious negative effects.
- Erik Hoel (part 2): Dreaming, AI, and the Future of Education — Finally, how do we think about the consciousness and theory of mind of AI models? Erik Hoel is a consciousness researcher and the writer of The Intrinsic Perspective, and this was a delightful look at how AI and humans dream.
In addition to that robust series, we had several excellent episodes that weren’t about AI (shocking!). Four favorites:
- Why quitters are heroes with “Quit” author Annie Duke — Annie Duke is amazing, thoughtful, and translates pathbreaking research on the psychologies of decision-making into everyday life. I found this to be a profoundly interesting discussion on not when to keep going, but when should we stop.
- Fertility Rules from wildfire sperm death and microplastics to the potential of AI w/ Leslie Schrock — Driven by personal experience, Leslie Schrock has made investing in fertility solutions her life’s work. It’s a far more expansive and complicated field than I could have ever known, and this episode gets at just some of the complexities of birthing new life.
- The Science of Survival: Adapting Human Life for Other Planets — Christopher Mason is a professor with a bold view: we need to genetically engineer humans to be adaptable to outer space if we are ever going to live off-planet. This is our most speculative show of the year, but it was a fun one.
- Navigating the Crossroads: Technology, Democracy, and National Security with Miles Taylor — finally, formerly anonymous Trump official Miles Taylor has made it his mission to protect American democracy from demagogues. This is an incisive look at the challenges of keeping America the leading nation of the world.
Best of “Securities” 2023
AI was a recurring theme this year, as was international relations and the rise (and more specifically, falls) of the United Kingdom and Canada. But if there was one thread tying “Securities” together this year, it was employment, jobs and the meaning of work. From the strikes in Hollywood and the rise of generative AI to tighter immigration restrictions across the developed world and the increasingly nihilistic work life of most jobs, we sit at a critical juncture on how work should be valorized within our identities as people and countries.
My five favorite “Securities” columns from the past year:
- "Existential Engineer" — my favorite piece of the year departs from the blind optimism of most of the tech industry to pursue a more deliberate path toward an existentialism of engineering and why building things matter to us not just materially, but spiritually as well.
- "Garrulous Guerrilla" — the essential thesis on AI. I argue that the small fraction of creatives who do truly original work will survive generative AI, but that very few creatives are equipped to do original work. The implication is that millions of people will eventually lose their jobs in the years ahead.
- “Professional Prerogatives” — the jobs that will be protected from AI are those professions that hold the power to stop it cold. First up on that list will be doctors, who have already rebranded artificial intelligence as “augmented intelligence” in their pursuit of human autonomy. They will succeed, but other professions with weaker organizing skills will likely lose.
- “Brainwash Departures” — governments around the world are putting in place more and more restrictions on workers with expertise, essentially arguing that the thoughts of some workers are so important, that they are a national resource that must be secured. It’s a chilling new pattern, and one that should be aggressive fought against.
- “Striking Employment” — the superstar effect is decently studied in economics, but its effects have expanded to many more labor markets and even to many industries as well. The best are taking a greater share of the returns, and that’s transforming the economics for everybody else.
I’ll highlight two more pieces from our guest writers this year. Our summer associate Ken Bui’s best piece this summer was “IP, IP, Betray,” and our freelance writer Michael Magnani did a great piece on international organizations in “Loose BRICS.” Thank you both for joining “Securities” this year.
- Grace attended the epicenter of AI that is NeurIPS last week in New Orleans, where our portfolio company Together’s academic partner Dan Fu won best paper for FlashFFTConv alongside his co-authors Hermann Kumbond and others. Check out the paper as well as his more accessible tweet thread. “We show that FlashFFTConv improves quality under a fixed compute budget, enables longer-sequence models, and improves the efficiency of long convolutions.”
- Alex Nguyen recommends the Apple TV+ series Drops of God, which pits the daughter of a famed wine oenologist against a competitor for the inheritance of tens of thousands of bottles of rarefied wine. It’s a French adaptation of a Japanese manga series, demonstrating both the globalized nature of wine and also media.
- Brooding and dark, I found Carolyn Dever’s essay on “How to Lose a Library” to be prophetic and philosophical. It’s about the cyberattack that has completely knocked out the British Library for weeks now. “How ironic that the most quaintly analog form of research possible, using physical books in a physical library, has been devastated by the hijacking of a digital system. I am experiencing this irony as especially bitter this morning, having arrived at desk 1086 with my list of tasks, hoping against hope that the crisis had resolved. It hadn’t. I hope it will someday soon.”
- Shaq Vayda enjoyed Siddhartha Mukherjee’s latest piece in The New Yorker on “All the Carcinogens We Cannot See.” “This is a chilling duality of cancer: each individual cancer comes from a single cell, and yet each cancer contains thousands of clones evolving in time and space. Treating or curing cancer involves tackling this incredible degree of genetic diversity. It’s a clone war.”
- Finally, our scientist-in-residence Sam Arbesman has a couple of fun pieces to note. An interesting fact about a Wikipedia article, why the oldest people are not as old as we think, and finally, the age-old question about the movie Home Alone — is the McCallister family rich? The New York Times investigates.
That’s it, folks. Have questions, comments, or ideas? This newsletter is sent from my email, so you can just click reply.