3 ms·
I believe there is more to it than just problematic or “almost correct” code examples after 7 retries with ELI5 explanations. But part of me believes this is a
by atmosx 3y ago
I believe there is more to it than just problematic or “almost correct” code examples after 7 retries with ELI5 explanations.
But part of me believes this is another “self-driving car” thing where in 10 years we’ll still be in exact same place.
Today I saw an interview from a Greek YT economist who was asked about BTC. His answer was “it’s been 10+ years that has bene around. We know by now that is not new or groundbreaking in any substantial way” and I agree. There are many tech that “if this and that” but we never seem to reach “that” point.
- aydyn 3y agoBut self driving has improved. Maybe not as fast as people would like but its clearly better than it was 5 years ago.
- lm28469 3y agoYou can't tell people you're "24 month from full self driving in every driving condition" every 12 months for a decade an expect people to continue drinking the koolaid It's the same for "AI", it's happening in cycle since the 60s, nobody learns a damn thing, it's borderline a pump and dump scheme at that point
- CamperBob2 3y agoYou're making a terrible mistake.
- lm28469 3y agoI can come back to this discussion in 10 years and we still won't have self driving cars. I wouldn't even be surprised if personal cars are banned entirely before we reach self driving cars
- CamperBob2 3y agoWe're not talking about self-driving cars. That's not a technical problem, it's a human problem.
- DennisP 3y agoNo we can't. Instead we can expect people to start dismissing it, despite the fact that it keeps steadily improving, so when it finally gets good enough it takes everybody by surprise.
- ben_w 3y ago> But part of me believes this is another “self-driving car” thing where in 10 years we’ll still be in exact same place. I want us to be still having these discussions in 2034, because that would mean I'm still capable of gainful employment for the next ten years. But a mere 3 years ago, what ChatGPT is doing now — even with the mistakes — was SciFi level of natural language processing.