I’m failing to see what’s so advanced about LLM usage in the big labs. If it was so advanced, they’d show it so it drives demand/pricing/IPO but I don’t see anything that local models can’t do. Businesses haven’t adopted the tech because it takes time/thought/economic pressure to innovate.
What does it mean to “genuinely” contribute to mathematics? Because the definition you use is load-bearing ;) and it might be different from that of others.
This seems like a step towards their plan for building a platform in competition of Google/Apple. Once the generative UIs are polished and more useful than individual apps, their hardware can now ship a device which circumvents the app store moat.
I believe every individual and organization should understand the tradeoffs between the two and have a backup plan. The backup is typically unnecessary assuming stable times but when incentives change, everyone should be vigilant for rug pulls. As an individual, I can calmly build a self-hosted inference stack based on EOL crypto mining accelerators but when building for clients, support and replacements becomes more important than budget and novelty.
Do you think car manufacturers hire people who can program or even care to program this stuff into the tracking? Knowing some of the people who those tasks get outsourced to, the answer is no.
User drives through grass which the sensors detect as hard collision. Gettem, NexusLexus!
Do you think insurers care to differentiate between those things? Looking at the equivalent behavior from UnitedHealth and their auto claim rejection behavior, my guess is also no.
Any data collected about you can be and will be used against you.
reply