Wild, Wild Times

Interesting. Anthropic just launched Claude Duet in research preview, and I’d been wondering when Claude would get into the agent space. Grok Bot, Meta’s Muse, and OpenAI’s Dots seem to be pretty popular, and all indications point to a fair amount of uptake for these agents, even if it’s mostly the early adopter crowd for now.

I think agents are here to stay. Personal agents are here to stay. We’re heading into a world where these agents will be to normies and tech-savvy people alike what Jarvis and Friday were to Iron Man. An always-on, persistent agent running around the cybersphere, doing things for us.

To be fair, we’re not quite at Jarvis yet. Duet still shows you what it’s about to do and waits for your okay before it sends a message, makes a payment, or signs something, and some things, like trading securities or moving money between accounts, are off-limits even if you say yes. It’s built mainly for work, so personal errands like booking tickets aren’t its strong suit yet. So for now it’s more careful assistant than Jarvis. But give it time.

What blows my mind is that all of this has happened in the last year and a half. And what’s even more mind-blowing is that ChatGPT only launched in November 2022. All the progress we’ve seen with LLMs has happened in just under four years. The pace at which these capabilities are improving and compounding is ridiculous. I’ll admit I’m a little historically uninformed here, but I can’t think of another technology that’s improved at such a rapid clip.

What still bothers me, and I keep saying this over and over again, is that this is magic, and people seem to take it for granted.

Wild, wild times.