Sam Altman talks GPT-4o and Predicts the Future of AI
EP 104 of The Logan Bartlett Show: Untold stories from tech's inner circle
Just hours after OpenAI's GPT-4o launch, I sat down with Sam Altman to discuss the new release, and Sam shared his biggest predictions for the future of AI. Some key takeaways...
Click here to view the episode transcript | Watch on Youtube | Listen on Spotify or Apple Podcasts
✉️ Episode Memo
→ AGI is only 1-2 ideas away
Sam initially believed there would be a clear moment when AGI emerged. Now, he thinks it will be a gradual progression, with no consensus on when it happens. It’s not GPT-4 and probably won’t be GPT-5, but we’re probably only a couple of breakthroughs and a bit more scale away from AGI.
→ Should AI be regulated?
Sam predicts that AI will pose significant catastrophic risks to the world at some point, and when this happens, regulatory oversight will be helpful. However, he believes it would be a shame to regulate it prematurely because of the tremendous upsides of the technology. He also views safety as a spectrum rather than a binary state, comparing it to the risks of air travel: while the probability of safety is high, the risks can never be fully eliminated.
→ Why Sam doesn’t think AI girlfriends will take off
Humans are inherently wired to care deeply about other humans. Our obsession with other people is too strong, and we just aren’t as emotionally invested in AI relationships. This is why society creates conspiracy theories about Sam but not ChatGPT. For example, society will probably never choose to watch AIs play soccer over real athletes.
→ Why humanoid robots are so exciting
Sam believes we should design a future where humans and AIs can work together seamlessly. He’s particularly excited about humanoid robots because the world is already designed for humans, and he doesn’t want that to change. He envisions AIs communicating with us in human-optimized language and even interacting with each other in the same way.
→ The importance of separating you & your AI assistant
In five years, it will be valuable to know whether you're communicating with a person or their AI assistant. Sam believes in keeping this separation distinct, so people know if a person or their AI assistant is handling tasks and messages. Additionally, he personally prefers viewing the AI assistant as a separate entity rather than an extension of himself.
These points only scratch the surface of Sam's vision. Listen to the full conversation for more insights.
⭐ Trailer
📱 Follow The Show!
If this was helpful to you, please consider forwarding!