Same with prompts, most attempts seem to be fidgeting with the models till they get your intend right, which is also a matter of hill-climbing, subtle mutation, and so on.

If I were to clarify anything from the article, I'd probably say that I'd rather do the factorisation of programming roles by how long they already existed. If someone is an AI engineer and his work only became relevant a month ago, very probably it will be obsolete in another month. If they do the same thing for the past 10 years, changes are that their skills would be useful for another 10 years to come.

Of course not, all these sloppers are doing is training the models so at the eyes of management they are good enough for a replacement. The ones who stay will have 10x more work.

Although this may be more relevant to replacing AI researchers, not AI engineers...

(Submitted as https://news.ycombinator.com/item?id=48380643 )

Most training is now actually inference, not directly gradient descent. Reinforcement learning requires the generation of lots of 'rollouts' that are then compared with each other via an algorithm like GRPO. Or they might be compared using a critic model - AI judging AI and causing it to self improve. Generating a rollout means inference. And there's lots of data cleaning by older models. This has been called in the past 'textbook' or 'curriculum' learning, not sure what it's called now. But AI is also used for things like data/document labelling, transcription of videos, detection of images/videos with watermarks or subtitles, elimination of content that shouldn't be in the dataset, creation of new content that should and so on.

AI has proven capable of some routine work, like brute-force optimizing GPU kernels or doing hyperparameter sweeps.

Obviously, researchers are all using coding agents too.

So that's a few ways AI is self-improving. But there are lots of other ways in which even frontier models are still beaten by human researchers. Experiments in closing the loop have failed. For instance, people have tried giving the latest models access to some GPUs and an old version of an AI codebase that was recently optimized by human researchers (a NanoChat speed run goal, I believe). Could the models match the performance of the AI researchers? Nope. They only got 10% as far as the humans did, mostly because their approach was uninspired. They wasted a lot of time and budget doing low-IQ stuff like hyperparameter tuning. The humans had many other tactics like studying the research literature and inventing new algorithms that the models didn't even attempt.

The bottleneck is therefore currently the level of insight and inspiration the models are capable of. I've also seen this in my own work. I come up with an idea I think is novel and see if I can get a frontier model to reach the same idea. It never works without questions so leading it's more or less pointless.

It's very unclear why AI struggles so much with innovation yet can invent new songs, poems etc without apparent difficulty. Obvious answers like "it's not in the training set" don't feel right to me, the issue is deeper.

In fact, I believe that the most cost effective way is a collab of human+agent. Ie giving the agent direction as it goes along with the plan I can cut the thinking while keeping the speed. Basically helping the agent going from a breadth first search into a guided depth first one which is much more token efficient.

Additionally, humans have long term memory and knowledge of the context around your codebase. Agents do not, and while you can fit a lot in 1M context window, once you fill that the quality goes down considerably.

The more interesting question is: Has the ease of replacibility increased because of AI?

These are things I've come to expect from bots, clueless journalists, clueless juniors, clueless expert beginners and clueless members of the professional managerial class but almost never from experienced software engineers.

To be fair, seasoned software engineers always seem to get shouted down online by the former group which is louder and more numerous so you could argue that we "lost" the argument.

Meanwhile big tech's vibe coded monstrosities are increasingly exploding all around us in ever more humiliating ways while the humans who had this tech rammed down their throats get thrown under the bus.

This undeserved halo effect over AI is maintained in order to keep the needle from pricking the ginormous stock market bubble that hinges upon the religious belief in the lie AI Will Replace Us All Soon.