@pmddomingos That's right. Newton came up with convergence centuries ago. It works very well with quantitative data, but not so much with qualitative, which is a judgment call on what 'improvement' means.
@pwlot Save this post. I doubt it's going to age well. If you're right, put everything in on it and decide to become wealthy. It can only be transformative if it's sustainable. It's not.
@erikbryn Not perfect? It's the inverse of a good metric. More LOC == more liability. I'm guessing your background isn't SWE. Besides, the second you turn a metric into a target, the metric is useless.
@TMTLongShort This guy isn't ready to walk past a used car lot or vacuum cleaner salesman. They've been grifting and moving goal posts with consistency, but you can DEFINITELY believe them now. 🙄
@PeterDiamandis That's the optimists read. The practical reality for mere mortals is that limited time is going to incentivize the shortest path from A to B. It's not rocket science, and for most people, it will weaken their skills, not improve them. Should we have easier lives or be better?
@mark_k Nonsense. In very limited domains a convergence can be achieved based on some metric, but most knowledge domains are qualitative, not quantitative. The definition of improvement itself is a judgment call.
@Prathkum What is really implied is a turbo charge effect where the model rapidly improves itself into expanding capabilities and features without needing human input. By that definition, we aren't there.
@AlfinCodes It's not even 'AI as it stands'. We live in the most dynamic period in SWE. Your amazing workflow today can be turned upside down or made obsolete in an instant, and if a tool that you become dependent on becomes obsolete, you will move below someone who can thrive without it.
@VigilantFox These people are FOS. They have no clue, and they'd lie if the truth would sound better. Half the alarmists are actually working for the tech companies. Fear sells.
OpenAI is absolutely cooked. This is loser language. You can’t be four years into the bubble saying “yeah our customers have a huge issue with how expensive our business is.” You just raised $122 billion! You can’t say shit like this!
https://t.co/1uKEEpSS03
@Briankeating Chess isn't "solved". The programs are good enough that humans can't generally know the 'why' behind a move, but that doesn't mean the moves are perfect. They're just better than any person.
@Mrgunsngear@NancyMace That's the tragedy. If you allow a break for sub groups, it weakens the voting bloc against taxes. That's a game they've been playing for decades. That's why you have a 401k tied to your employer and the market. They give you a few crumbs so you won't vote against investors.