hackathonguy

I'm very curious whether a) anecdotally, anyone has encountered a real enterprise cost-cutting effort focused on LLM APIs and b) empirically, whether anyone has done any research on price elasticity in LLMs of different performance scales.

So far, my experience has been that it's just too early for most people / applications to worry about cost - at most, I've seen AI to be accountable for 10% of cloud costs. But very curious if others have other experiences.

I love that the base model starts with 16GB of RAM here. The value of these computers is incredible - I purchased a Macbook Pro in 2021 and it's still powering through every task I throw at it. Before Apple started making their own chips, I felt like I had to upgrade every 2-3 years to prevent my laptop from becoming a hurdle in completing every day tasks (remember when tab management was a thing?). Really happy with these machines.

I'm curious if it might be possible that an AI "civilization", similar to the one proposed by Altera, could end up being a better paradigm for AGI than a single LLM, if a workable reward system for the entire civilization was put in place. Meaning, suppose this AI civilization was striving to maximize [scientific_output] or [code_quality] or any other eval, similar to how modern countries try to maximize GDP - would that provide better results than a single AI agent working towards that goal?

In Jewish law there is a paradigm called "this one is benefitting and this one is not losing". Essentially, the idea is that if you're enjoying something belonging to someone else, that doesn't constitute stealing as long as they don't stand to lose anything owing to your usage.

This applies to lamplight, and interestingly, can be applied to piracy as well: if you can honestly say you wouldn't have purchased the movie/software/music anyway, downloading it is not stealing.

As others have mentioned in the comments, this idea doesn't work for wifi: performance is inversely correlated with connected devices, making using your neighbor's wifi a case of "this one is benefitting and this one is losing".

Wondering if it's just me: I bought an M1 MacBook Air with 8GB of RAM, based on reports saying that RAM management is much more effective than with Intel CPUs. Ever since, I'm really struggling with load management. I typically have Chrome open with 20-30 tabs, plus a couple of Electron apps (Notion, Slack, Google Calendar), and my MacBook frequently slows down to a crawl or gets stuck entirely. Simple commands - "close all tabs" in Chrome - can take 50 seconds to execute. Is anyone else experiencing the same?