Special Intelligence
Special Intelligence is plenty ⓧ Automating software engineering ⓧ Superpowers in AI
👋 On time for your weekend: a round-up of this week’s remarkable stories at the intersection of technology, business, design, and culture. Three reads and three listens; no fluff, just stuff ⚡
We are drowning in information, while starving for wisdom. The world henceforth will be run by synthesizers, people able to put together the right information at the right time, think critically about it, and make important choices wisely —E.O. Wilson
📚 Reading
LLMs have special intelligence, not general, and that's plenty.
LLMs have the distilled knowledge of everything we've produced that creates some amalgam which is Specially Intelligent. It is able to connect the insights from any domain with any other and be able to give us a prediction of what we might say if we had thought to connect those tokens together. Sometimes coherent, sometimes not. It is also not General. Emphatically not.
Rohit Krishnan—Strange Loop Canon | 11 minutes
Automating software engineering:
There is a lot of work not just on the AI part but also the UI/UX part. How does a human provide oversight? What are they looking at? How do they nudge the AI down a different path? How do they debug what went wrong? It is very likely that we will have to change up the code editor, substantially […] In any case, software engineering is on track to change substantially
Andrej Karpathy—X | 2 minutes
Which AI should I use? Superpowers and the State of Play:
It suggests an alternative (and I think maybe more successful) model of integrating LLMs into organizations than chatbots and custom applications: treating AI like people who can be added to standard organizations and teams. You can imagine other agents in charge of marketing, or research, or analysis. These sorts of agents are going to be another big theme of the next set of AI releases.
Ethan Mollick—One Useful Thing | 11 minutes
🎧 Listening
AI's Influence on the Economy:
It's about there's things that humans are not than AI at, but are better relative to the things they're really bad at. We will allocate AI toward the things that AI is most better than humans at in a relative sense. Whereas the things that humans are only a little bit worse than AI at will be things that humans do […] There's a fundamental reason to be optimistic, which is that AI being better than us at stuff doesn't mean that AI will be the one doing the stuff.
Noah Smith—Econ 102 | 74 minutes
Platform Shifts:
What's the plan? That Apple becomes 20% of global GDP? […] Google's free, Apple just builds the best product, and so these companies are never going to get any smaller, and there is no way to launch a phone competitor today. And that's not because they're doing stuff that might be illegal, but it's just because they've gained such an advantage by being so good at what they do that we can't stop them.
Alex Schleifer—People vs Algorithms | 78 minutes
The happiness and pain of product management:
Ideally, things that are best for the customer, there's high overlap with that with things that are best for the business, but not always. Figuring out some principles that help guide those conflicts can be really, really helpful. At Thumbtack, we had principles about which sides of the marketplace we wanted to serve in which order and when we serve Thumbtack. Saying […] is the easy thing to say. Actually doing it in action is a very different thing.
Noam Lovinsky—Lenny’s Podcast | 70 minutes
💎 Timeless
1️⃣ year ago—GPT-4 is founding an Empire of Imagination
2️⃣ years ago—The Arc of the Practical Creator
3️⃣ years ago—Innoveracy: Misunderstanding Innovation