0xd1r

How does it do so? Just DOM manipulation, viewport scanning or something of the sort?

Can it perform DOM manipilation as well, like fill forms or would the LLM response need to be structured for each specific site to use it on? And would an LLM be able to perform such a task?

> As part of our IMO work, we also experimented with a natural language reasoning system, built upon Gemini and our latest research to enable advanced problem-solving skills. This system doesn’t require the problems to be translated into a formal language and could be combined with other AI systems. We also tested this approach on this year’s IMO problems and the results showed great promise.

Wonder what "great promise" entails. Because it's hard to imagine Gemini and other transformer-based models solving these problems with reasonable accuracy, as there is no elimination of hallucination. At least in the generally available products.

I suppose they are using keras because of its simplicity. But I agree, things have really moved on from keras.

One of the best passion projects I've seen. Love that there's so many one-click deploy options.

Indeed, quite lovely to see. Wish someone did a piece on 'Metro Gali', where there are people like this for cameras.

This reiterates the same fallacious argument large corporations use to put one over open-source. And since it sounds plausible to the layman, it is parroted by them. Quite shocking to see another example out and about.

Wasn't 1965, part of ALGOL W, but somewhere in the previous millenium, in a language meant for object-oriented tax collection.

I'm hoping that this is a commit that is reverted back, after some amount of backlash from the community, if at all this was intentional.

Else, it evokes the sentiment of riskynacho on the comment, "It's almost like DuckDuckGo wants to eliminate itself from being a safe useful alternative in the search engine competition."