To this day I do not get why Intel doesn't just offer massive memory options for their cards. Just charge what it costs to add the extra memory, no upcharge, and they will never be able to keep up with demand. Cheap VRAM is enough to justify a lot of open source investment into challenging CUDA.
That was the main reason for the big hype around Memristors 15 years ago. High density, high speed persistent memory to completely remove the need for hdd/ssds, potentially even removing the need for external memory altogether. So frustrating that it still seems like we're a long ways from that becoming reality. There's some renewed interest in Memristors as they can simulate neural network connections in models, so maybe the funding will return for it.
Religion relies on assumptions that cannot be proven physically, secularism just avoids those assumptions (I say this as a Christian who finds secular biblical scholarship fascinating and helpful).
Even without encoding, as long as decoding is supported for AV2, streaming sites like Youtube can always transcode uploads. The encoder on mobile hardware is more of a nice bonus as long as we have an AV1 encoder available in the meantime.
My hope is that hardware improvements (better node densities every 2-3 years, better designs, etc) will pick up the majority of the savings for these companies in the future, assuming LLM performance starts to taper off with diminishing returns.
You learn to know who are the lazy ones, and at that point you can politely always respond with a, "what were you able to find on this?". You can repeat this ad infinitum since at that point, they're just being lazy and disrespectful of your time. They eventually give up going to you because they know they won't get you to do their work for them.
My stomach churns when I see an e-mail obviously written by AI. It just comes off so disingenuous and disrespectful, as if they couldn't be bothered to respond to me themselves but instead had a machine do it. Even worse, it's usually very verbose, wasting my time further to have to read through all the slop.
It's funny, when we were acquired they started moving us to Github actions but it seems that maybe we should stay on our old crusty self-hosted Jenkins setup...
Certainly you realize that these comments exist for more than a single person right? You expect potentially hundreds of viewers to each burn through AI tokens instead of just getting a direct and relevant answer here? This has the same vibe as the old forum posts where the only response was a "google it".
That used to be the case, but I would argue that C#'s ecosystem is just as competitive now. For our backend tech stack at work, all of our libraries used are open source nugets. Back 15 years ago, paid nugets were definitely more standard.
As a percentage of their income? Because that's the only number I care about. You don't get to hoard wealth off the backs of tens of thousands of workers and then act like paying a smaller percentage is some good deed being done. The more one benefits from society (and billionaires depend most on the financial security and infrastructure setup by society), the more one needs to pay back into the system they gained their wealth from.
The first step we need to take is to invest in the IRS. Every dollar invested in the IRS returns between $5-9. Couple that with fines that offset the cost of auditing, and "hidden wealth" becomes a liability too expensive for people to bother with.
Obviously there's an ideal middle ground, but what LLMs do is allow free transfer of knowledge while still (mostly) preserving the protections that copyright should be protecting. For example, I can have an LLM give me the entire plot of a book (which is fine), but it won't spit out an exact copy of the book.
That's more an argument for why you can't just use LLMs as a source of truth. Conveniently, LLMs like ChatGPT do often cite their sources, especially if you prompt them to.
Yeah this is the main argument I have for AWS. Is it the best tech-wise? Arguably not, but they have real support you can contact on a moment's notice, they'll even reach out ahead of time if they suspect something is seriously wrong.
This might work but only if the evaluations are done through a trusted third party entity where none of the money ever reaches the company you're submitting to.
Reminds me of the paper launches NVidia/Intel/AMD sometimes do where they announce some amazing tech (such as the old Titan GPUs) that placed their hardware at the top of the benchmarks, but with basically zero actual stock available.
I think CEOs are so drunk on the shareholder buzz for AI that they think this is also what the customer wants. I love integrating AI into products, but only in a way that is seamless. For example, I did my taxes recently and there was a button to upload my tax pdfs to do a best effort auto-fill of some forms on the tax website. No mention of AI even though AI was almost certainly used for that button, just a simple plain button.
To set an example, as a deterrent to larger operations. If they go after even the smallest impact counterfeiters, it leaves no room for a plausible level at which an operation can safely run under, no matter how small.
Same. I don't care about the online connectivity or whatever, I just print a few personal things every month so the convenience and reliability far outweigh any cons for me.
My favorite is when Claude will build a completely new application to load and inspect a .dll file using reflection instead of just googling the library's interfaces.