Isn't this one mimicking the typesafe ai horror website ?

That's unless the code produced in the future is much more complex than today's.

Sure, but it would be actively bad to make the code more complex simply because we have machinery that helps us deal with the complexity. A big part of how people assess the models' coding capability is whether they create needless, incidental complexity.

That's like saying it'd be actively bad to make the code more resource intensive simply because we have machinery that helps us deal with the extra requirements. And as we know as computers got more powerful code didn't get lighter. If it can, it will.

Using more resources than necessary is actually waste.

That's my point

I'm confused. You started out with "That's like saying", but you appear to agree with me completely.

What I mean is we have evidence that if it's easy and possible people will do it. There's no need to predict anything. That it is bad is besides the fact, although obviously less performant software is worse, its easier to create.

Upthread we started with

> Compute has already lost value for me. Six months ago I thought you needed a 1T+ model to be useful coding. Now I am able to get by just fine with a 27b model.

and I was explaining why the complexity question is important in that context. Specifically, because it's part of why people judge the 27b model to be adequate compared to the 1T+ one.

Yes, lots of people are slopping out low-quality things with powerful models because they don't know how software quality works and expect magic. I think we are talking at cross purposes but have no real disagreement.

Assuming it’s all going to be vibe coded garbage, yeah it will be much more complex. Like a toddler writing a symphony.

The risk is they are a magnet for many more talented ML scientists to leave google.

> a magnet for many more talented ML scientists to leave google.

Also to leave Meta, Amazon, Microsoft and everywhere else.

There would be more people who wouldn't join Google, but would love to do this instead.

Daily sports is the way, purges the body, rests the mind.

But not too late, otherwise you don’t sleep well.

Yes, full ft or lora https://github.com/NVIDIA-NeMo/Automodel/blob/main/docs/guid...

Yes, all my emails gyer sorted out by a finetuned gemma. There are turned into images passes to the model, as multimodal is so practical.

Reminded me of the recent, and excellent, Canadian tv series Empathy, with the main character is found in a garbage can by his adoptive parents.

AlphaEvolve couples map-elites with LLMs. It's an key step in machine learning, in the vein of DQN for reinforcement learning.

AE brings diversity from the genetic algorithms community to large scale optmized deep learning and RL models.

It is a mandatory step for moving forward. The approach is clean and simple, while generic.

The only caveats is the per optimization problem definition of the map élites dimensions. But surely, this will get tackled somehow over the next few years.

If you don't know about map-elites, go look up Jean-Baptiste Mouret' s work and talks, it's both very interesting and universal.

Slides, publications and tech reports, very handy for figures !

Most software developers that I know spend only a fraction of time on that, if at all.

Generating diagrams is much more common than generating "images". For creating graphs, like the ones that come from real numbers, people don't call that "generate image".

It's useful when using prism, and for exploratory research & code.