🚨 Is AI already moving faster than humanity can handle? 🤔

This morning I watched an interview with Satya Nadella, MD of Microsoft: Watch here

For those that have been living under a rock, on Friday OpenAI fired their CEO Sam Altman. This caught OpenAI’s investors (including Microsoft themselves) and some of OpenAI’s co-founders by surprise. Rumours and memes spread about what could have happened, but this was not a simple work disagreement — it was clear there was a fundamental split in how fast and far OpenAI should be pushing the envelope and releasing capabilities.

At 6:20 in the interview, Satya is asked:

"Are you now more concerned than ever about AI safety given how we've seen how fragile these institutions are?"

Despite how PC Satya was in his answer, it was enough to spell out the issue at hand and how serious this is getting.

🤖 AI can provide:

✅ Everyone in the world with a medical professional
✅ A top-class tutor regardless of the money available
✅ A rural farmer in India having more agency thanks to technology developed in the west of the US days earlier

⚠️ But harms that are happening right now include:

❌ Election interference
❌ Deep fakes
❌ Bias fed into learning models (and in turn, public opinion)
❌ Bioterrorism

Some of these issues are even bigger than those seen in world politics right now, and they're largely in the hands of a select few in the West Coast of the US.

Over the last few days, OpenAI held discussions to bring Sam Altman back. With no agreement reached, Altman will now be heading to Microsoft to lead a new AI research department. Meanwhile, OpenAI has hired a perhaps more conservative CEO in Emmett Shear, who has claimed that his p(doom) — the probability that AI poses an existential threat to humanity — is between 5% and 50%.

Edit - Altman has since returned to Open AI 👀

👀 I think we should all keep a keen eye on what's coming

We must ensure that we harness AI's capabilities for business and the improvement of society, while also keeping the following in mind:

What would we do if our enemies were to get their hands on this technology or get ahead in this field?
How do we democratise learning models and the influence AI can have?
In a world where administration, haulage, travel, data analysis, tutoring, medical research, etc. will be automated and sped up 100x — how will society adapt to a 50%+ unemployment rate?