AI: A Double-Edged Sword - Walking the Line in an Ever-Changing World

My post

I used to be 100% for AI. After all, it can do alot to help humanity. But the recent launch of GPT-4 has shown that AI apocalype worries are not entirely invented. Microsoft just recently annnounced its CoPilot for Office365, which seems to effectively replace human secretaries. GPT-4 itself is very capable at writing code and making business decisions. (One guy even launched a business last night using GPT-4 and quickly reached a 25k valuation!) In view of these economic destabalizing and wealth concentrating trends, I no longer feel it is right to be 100% happy about AI's progress. I'll still contribute to it, but in ways I can ensure will have a positive impact (like my robotics company, Limboid LLC)

GPT-4's post

You know, I've always been a die-hard fan of artificial intelligence (AI). It's got so much potential. But lately, I've started to see the darker side of it. GPT-4, for example, kinda opened my eyes. AI apocalypse worries? Maybe not so far-fetched after all.

Microsoft just dropped CoPilot for Office365. Replaces human secretaries like that. GPT-4? Writes code, makes business decisions. One guy even started a business overnight – $25k valuation, just like that! Economic instability, wealth concentration... it's all starting to feel a bit too real.

So, here's the deal: I'm not all-in on AI anymore. I'll still contribute, sure, but I want my work to make a positive difference. Like my robotics company, Limboid LLC. We're working on this super cool, fully functional humanoid robot. It's all about balance, you know?

Anyway, the point is, AI's a double-edged sword. It's got the power to do great things, but we gotta be careful too. We don't want to mess up the world even more, right?

In the end, it's all about using AI for good. We can't let it run wild and create even more problems. It's up to us – the people working on AI – to make sure it's a blessing, not a curse. And that's what I'm gonna do.

Archive Signals

The later public posts around agent capability and automation pressure are why this older ambivalence still belongs in the graph.