ten years from now(assuming it doesnt crash our economy in more ways than one) i would imagine the software will at least be comparable to an entry level programmer.
That’s a big assumption. Also, AI stopped progressing in the past year, which shows it’s already hit its ceiling.
Like the user above said, virtually all human code has been fed to the models, now there’s very little new original human code been put online and with more adoption of LLMs in coding will only exacerbate this.
It is a massive assumption, but i tend to operate on the belief that capitalism will fall soon and some decent system will appear, because otherwise all this is a moot point anyways. Without the destruction of capitalism well all be dead within 20 years, so i dont find it as entertaining to speculate about a future where we are still bound by things like our economy. As far as the progression of ai this last year, didnt claude just learn to lie? Thats a huge step in development, not because its good, but because it shows our models are progressing. They know what were aiming for and theyre willing to lie to get to that target, which does actually show the most basic level of problem solving. Thats… huge actually.
That’s a big assumption. Also, AI stopped progressing in the past year, which shows it’s already hit its ceiling.
Like the user above said, virtually all human code has been fed to the models, now there’s very little new original human code been put online and with more adoption of LLMs in coding will only exacerbate this.
It is a massive assumption, but i tend to operate on the belief that capitalism will fall soon and some decent system will appear, because otherwise all this is a moot point anyways. Without the destruction of capitalism well all be dead within 20 years, so i dont find it as entertaining to speculate about a future where we are still bound by things like our economy. As far as the progression of ai this last year, didnt claude just learn to lie? Thats a huge step in development, not because its good, but because it shows our models are progressing. They know what were aiming for and theyre willing to lie to get to that target, which does actually show the most basic level of problem solving. Thats… huge actually.