System Prompt
'AI' Manhattan Project ⓧ Distilling with 'AI' ⓧ Horseless Carriages (again)
👋 On time for your weekend: a round-up of this week’s remarkable stories at the intersection of technology, business, design, and culture. Three to read and three to listen to—no fluff, just stuff ⚡
There is absolutely no inevitability as long as there is a willingness to contemplate what is happening—Marshall McLuhan
📚 Reading
Energy suck:
The AI data center buildout seems wasteful to me because it’s duplicative: a bunch of companies all racing to acquire exactly the same thing. Imagine another timeline […] bringing investigators together into super-powered labs, funding the construction of national data centers, making them available to academic researchers — the pioneers who have, in our timeline, been effectively sidelined.
Robin Sloan | 4 minutes
What I've learned about writing AI apps so far:
Is what you're doing taking a large amount of text and asking the LLM to convert it into a smaller amount of text? Then it's probably going to be great at it. If you're asking it to convert into a roughly equal amount of text it will be so-so. If you're asking it to create more text than you gave it, forget about it.
Laurie Voss | 8 minutes
AI Horseless Carriages:
The modern software industry is built on the assumption that we need developers to act as middlemen between us and computers. They translate our desires into code and abstract it away from us behind simple, one-size-fits-all interfaces we can understand […] In the new world I don't need a middleman tell a computer what to do anymore. I just need to be able to write my own System Prompt, and writing System Prompts is easy!
Pete Koomen | 21 minutes
🎧 Listening
Programming with AI:
One of the things LLMs open up is the ability for people to start doing much more sophisticated things with their computers without having to learn Python and SQL and all of these sort of like the trivia, of these sort of different programming constructs […] A small newsroom that can't afford a software developer can do deep data reporting on the numbers coming out of their local government, things like that. That feels like a uniquely positive way to apply this technology.
Simon Willison—Half Stack Data Science | 50 minutes
OpenAI’s Social Network & Figma’s Potential IPO:
I see a company and they pitch themselves as an AI company, unless it's like an infrastructure company, you're not an AI company, you're a fintech company, or you're a legal tech company, or you're whatever you are. As corny as this sounds, AI is a technology, not a category. And right now I think you're seeing a lot of, I don't want to say funds get fleeced because it's like voluntary, but all these AI multiples for companies that are not AI companies.
Morgan Beller—More or Less | 48 minutes
Be Good at Two Things:
What we need to do is stop trying to actually create solutions around media. Instead we want to go back to basics, understand human psychology motivation and what causes people to act as they do do, think as they do, and decide as they do. And we've got to go back to the basics of human psychology and then once we've understood that we can decide which, if any, of these new media might be appropriate to solving the problem.
Rory Sutherland—One One One | 10 minutes
💎 Timeless
1️⃣ year ago—The Age of "De-Bossification"
2️⃣ years ago—There Is No A.I.
3️⃣ years ago—If we’re all so busy, why isn’t anything getting done?