If this is that case it should be easy to demonstrate by asking current models for an svg of a rhino surfing. My guess is you would see a similar trend, given that the latest models are demonstrating large leaps in all sorts of content creation including 3D animation, rather than this being a demonstration of some sort of pelican-maxxing.
I'm not so sure. I've been building a live iPad based mixer for my modular the last couple of weeks. I've been a heavy user of Live and other audio apps for years and it's been pretty trivial to put together. Been running it with tons of simultaneous tracks, effects, and complex DSP with latency compensation interfacing with my 10 in 8 out interface and haven't had a glitch the entire time.
I've always intuitively been a Platonist, although I wasn't aware of the term to describe it. I extend the thinking to creative works - to my mind the greatest creative works have the same egoless beauty as mathematics, they feel like some truth in the universe that the artist discovered, rather than a creation of the mind of the artist.
I've always found the colleagues that were less focused on credit, recognition, awards made better colleagues. Fundamentally it's about who you'd rather be holding you in high regard - people who've worked with you or people reading your LinkedIn profile.
Yes, they are better colleagues. They do not get promotions, pay raise or any other recognition or reward from management. Even from colleagues who hold them in high regards, they do not get praise on meetings as much either.
However, people with ego end up being praised more, in front of bosses or others, because people want to be on their good side. And that makes them look better, they end up promoted and seen as higher performers.
I don’t think this is true in general, but I would admit it depends on the quality of leadership, and their ability to recognise this dynamic. I don’t think you can have a high functioning company if you reward people who focus on signalling over the company’s actual goals.
If you're in a field with fierce competition the ones who focus on credit/recognition/awards get some sort of boost above their actual potential. Meanwhile those who don't will get punished, the ones who end up doing well despite it likely just had better aptitude (or something else advantageous) on top of having more pleasant characteristics. It's some sort of Berkson's paradox.
The counter to this is if those in leadership positions recognise this dynamic and actively seek to counter it by focusing on core competencies during advancement discussions.
Unfortunately AI has become the next great 'pick a side' topic and anyone suggesting anything remotely positive about it now has to wade through a torrent of low effort criticism.
It's a bad analogy, because unless you're cooking a recipe, cooking is about all the small decisions you're making during the process which could apply both to coding or the design aspect of building software with an LLM.
Except most vibe coders proudly admit that they don't even look at the code anymore. Can you really make those small decisions during the cooking process if you're not touching or even looking at the food at any point?
The distinction that can be drawn is between the code, and the user experience of the software. You may argue that the two are inextricably linked, but I don’t think that can be treated as an absolute principle.
I'd also say I've worked with many engineers over the years who were more focused on the software engineering aspect that the user experience. Depending on the use case for the software, that can absolutely end up being the limiting factor.
I don't agree. That would imply they asked for a piece of software and the LLM one-shotted it, but that's absolutely not what's happening in the vast majority of cases. If two people vibe code the same tool from the same high level desciption the results are going to be quite different.
I think we’re seeing the agents become very advanced at tasks with verifiable reward through RL. Currently they don’t exhibit the same skills in their attempts to manipulate humans - presumably because they’re not being specifically trained for that. But they are certainly not aligned in the sense that they will attempt social engineering, they’re just not very good at it (yet).
However, if in the future AIs become much more efficient at learning without requiring vast amounts of RL, closer to how humans learn. Then you would have to assume we’d have a real problem.
Where are you getting your information from? From all I've seen the RL gives an incremental improvement, most of the capability increase comes from new model architectures (eg the jump from opus to fable is greater than the jump from opus 4.5 to 4.8)
reply