Each technology can not stand alone. It takes a saw to make a hammer, and it takes a hammer to make a saw. And it takes both tools to make a computer, and in today’s factory, it takes a computer to make saws and hammers. —Kevin Kelly
👋 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⚡️
📚 Reading
AI Prompt Engineering Isn’t the Future:
[W]hat is a more enduring and adaptable skill that will keep enabling us to harness the potential of generative AI? It is problem formulation — the ability to identify, analyze, and delineate problems [..] The fact is, without a well-formulated problem, even the most sophisticated prompts will fall short. However, once a problem is clearly defined, the linguistics nuances of a prompt become tangential to the solution.
Oguz A. Acar—HBR | 8 minutes
How do you govern something that's part software and part people?
When it comes to AI models, especially as they get better than we could’ve imagined, it can become incredibly difficult to figure out what all they can do. But it’s hard, because we don’t have a good enough theory of what it actually is [..] Software, which is meant to be predictable and testable, are a bit like LLMs. And they’re also like humans, who aren’t as reliable but are incredibly versatile.
Rohit Krishnan—Strange Loop Canon | 18 minutes
Setting time on fire and the temptation of The Button:
AI will make a lot of previously useful tasks meaningless. It it will also remove the facade that previously disguised meaningless tasks. We may not have always known if our work mattered in the bigger picture, but in most organizations, the people in your part of the organizational structure felt that it did. With AI-generated work sent to other AIs to assess, that sense of meaning disappears.
Ethan Mollick—One Useful Thing | 10 minutes
🎧 Listening
Apple and Facebook, Crypto and VR, Open vs Closed Ecosystems:
This is open versus closed ecosystems, and you're talking to someone trying to build a permissionless decentralized protocol, so I very much understand the power of those systems. I want to see more of them in the world, but the revealed preference of consumers is that they want tightly integrated, really easy-to-use, delightful experiences [..] The only example of a mass-appeal consumer product that is an open-ish ecosystem is email.
Dan Romero—Moment of Zen | minutes
State of GPT:
Especially for the bigger models like GPT-4, you can just ask it, 'did you meet the assignment' and actually, GPT-4 knows very well that it did not meet the assignment [..] But without you prompting it, it doesn't know to revisit, so you have to make up for it in your prompts [..] recreating our System 2. System 1 is a fast, automatic process, and that corresponds with an LLM just sampling tokens; System 2 is the slower, more deliberate planning part of your brain.
Andrej Karpathy—Microsoft | 42 minutes
Marketing, Meaning, and the Bibs We Wear:
When someone shows up to tell a story — and it’s not just someone who’s trying to sell you a soft drink; it’s someone who’s trying to run for office or an economist who’s trying to get people to adopt her ideas — they are telling a very complicated story that involves everything from how is the typeface kerned to what institution is this person part of, to what do my friends think of them? [..] It is the stories we tell each other about who we are and where we belong.
Seth Godin—Conversations with Tyler | 54 minutes