They may add any number of tools, but that will add exactly zero logic and reasoning to models — the main "weakness" of everything stupid marketing calls "intelligence" these days.
> There are many things humans do that are very different from next token prediction, and those things we do all combine together to produce human level intelligence.
Exactly
LLMs didn't resolve knowledge representation problems. We still don't know how it's going in our brains, but at least we know, we may do internal symbolic knowledge representation and reasoning. LLMs don't. We need a kind of different math for ANNs, a new convolution but for text where layers extract features through the lexical analysis and ontology utilisation, and then train the network.
Well, I doubt that an ergonomy is the reason here. Going gym consistently benefits much more, rather than switch devices. I have never seen healthy programmers over 30 who avoided physical activity. But I saw how people from constant switching of ergonomic devices and clinics visits just in several months were able to forget about most of the pain in wrists and back just by consistent increase in their physical activity and adding some resistant training. Their back CT looks the almost same, but their lives are significantly improved.
I work on any devices and sit on various chairs for almost 25 years, and never experienced any significant pains. Just visiting gym and having some massages. In my case the most significant factor that affects back health is stress, but massages is what help to maintain it.