Funston Brief (Issue No. 21)
This month, our featured commentary discusses the opportunities of distributed machine learning projects. We also dive into the latest discussions on techno-optimism and SBF’s trial.
A global machine learning effort delivered the word "πoρφυρας," a name for a purple dye, in October, marking its first significant achievement in an unlikely domain - a recent effort utilizing the Herculaneum papyri, which were last read about two millennia ago. For those who study ancient history, this is an exciting period since it will be fascinating to examine old writings free of commentary. However, this distributed approach to AI and discovery holds broader lessons.
Emerging technologies frequently go through a phase of relative obscurity during which there is much opportunity for rapid advancement, but they also often hit some talent or money bottleneck that forces the participating enterprises to become more institutional. Because there is so much context to catch up on and because obtaining the necessary data or compute power has become more expensive, it is more difficult for someone to pop out of nowhere and make a significant impact on machine learning. Because of this, the papyri discovery is a solid predictor of a usable mental model in the future: industries downstream of AI are likely to have more opportunities and less rivalry. Perhaps learning ancient languages would be a better option than Pytorch if you want to be in the 99th percentile in a field that is evolving quickly.
ACTIONABLE INSIGHTS.
Techno-Optimist Manifesto: Co-founder of Andreessen Horowitz Marc Andreessen recently published a piece that garnered a lot of attention inside the tech community. It's a purposefully provocative manifesto, full of rhetorical clichés and one-sentence paragraphs, promoting optimism about the potential of technology to better the world. Ben Sixsmith delivered a measured response.
Key Insight: Ben Sixsmith notes that we are incredibly hesitant to give up modern conveniences (e.g., a flight without Wi-Fi or entertainment) and have a default mindset that may be described as vintage techno-optimism and future techno-skepticism. His essay discusses a problem that arises when technology simplifies complex tasks, eliminating the inner sense of accomplishment that comes from mastering something difficult. Marc’s article Ben’s response
Parallel Bets, Microsoft, and AI Strategies: Matthew Ball has a new piece on corporate strategy formulation in the “age of AI,” the inherent messiness of these strategies, and lessons from the past. Though his essay is narratively focused on “Big Tech,” it applies to any company evaluating AI - whether they focus on video, video games, music, enterprise productivity, automation, and more.
Key Insight: Parallel bet strategies have several core advantages, from increasing corporate optionality to covering strategic bases, maximizing learnings, and neutralizing (or at least moderating the risk of) potential competitors. But there are costs. More money tends to be spent, but each bet receives less funding than they might have under a more focused approach, which can constrain the would-be “winners.” Overseeing many bets can also lead to mixed signals on what is (and is not) the “right bet” and harm internal morale. Link
EAR TO THE GROUND.
Podcast Episode: Conversations with Tyler - Ambition, Art, and Evaluating Talent
Guests: Paul Graham
Takeaway:
Tyler Cowen and Y Combinator co-founder Paul Graham sat down at his home in the English countryside to discuss what areas of talent judgment his co-founder and wife Jessica Livingston is better at, whether young founders have gotten rarer, whether he still takes a dim view of solo founders, how to 2x ambition in the developed world, on the minute past which a Y Combinator interviewer is unlikely to change their mind, what YC learned after rejecting companies, how he got over his fear of flying, Florentine history, why almost all good artists are underrated, what's gone wrong in art, why new homes and neighborhoods are ugly, why he wants to visit the Dark Ages, why he's optimistic about Britain and San Fransisco, the challenges of regulating AI, whether we're underinvesting in high-cost interruption activities, walking, soundproofing, fame, and more.
“The earlier you’re judging start-ups, the more you’re just judging the founders. It’s like location for real estate [...] Determination is hard to judge by looking at the person. The way you would judge determination - because people act determined - [...] is not so much from talking to them as from asking them stories about things that have happened to them.”
- Paul Graham, Co-Founder of Y Combinator
1 million
Microsoft has over a million paying users for its LLM coding assistant, Github Copilot. Link
PROFILE.
During the FTX saga and Sam Bankman-Fried’s trial, Michael Lewis and his (more than) timely book on SBF - Going Infinite: The Rise and Fall of a new Tycoon sprung into the limelight. Lewis found less evidence than one would have expected, both before and after FTX's issues were made public, for someone who had essentially unrestricted access to the main players in what ended up being a rather major fraud trial. In his defense, Lewis didn't anticipate controversy and bankruptcy, and he most surely didn't anticipate having to decide what to do if his hero unexpectedly turned out to be a villain. Samanth Subramanian has a profile of Michael Lewis in The Guardian. And Christopher Beam has one in the New York Times. Enjoy the read!
“Michael finds these characters, falls in love with them a little bit, and makes them into heroes. The radical thing here is to withhold judgment.” - Former colleague of Michael Lewis.
ON MY RADAR.
The Silicon Valley elite wants to build an $800 million city from scratch. Link