101–110 of 1001 posts
Companies have hijacked the open source concept to mean downloadable blob and we follow them as I see in the comments. It’s a real shame.
But I agree, it's a real shame.
Companies have hijacked the open source concept to mean downloadable blob and we follow them as I see in the comments. It’s a real shame.
Better a downloadable blob than a non-downloadable one
A truly open model has open code that gathers pre-training data, open pre-training data, open RLHF data, open RLAIF data generated from its open constitution and so on.
The binary blob is the last thing I'd want - as a heavy user of LLMs I'm actually more interested in the detail of what all training data is in full, than I am the binary blob.
Earlier quoted context omitted.
And Anthropic not even in the top 10 ...
I keep hearing about Claude's impressive coding skills (compared to its benches) yet, not evident for me (I use the web version, not cline). Compared to 4o it's not that great.
In our small and humble internal evals it regularly beats any other frontier models on some tasks. The shape of capability is really not intuitive/1 dimensional
Grok has gotten to the top of one benchmark: https://x.com/lmarena_ai/status/1891706264800936307 It's been said before but it is great news for consumers that there's so much competition in the LLM space. If it's hard for any one player to get daylight between them & the 2nd best alternative, hopefully that means one monopolistic firm isn't going to be sucking up all the value created by these things
I've spent the last hour testing it and I'm blown away. And this is coming from a very hardcore user of OpenAI/Claude products on a daily basis. It passed every goofy test I have for writing articles which involves trying to surface arcane obscure details. (it certainly means however they are scraping the Web they are doing a good job here) It made the database code I wrote over the last week with o3/o1/GPT4o/Claude3…
Earlier quoted context omitted.
Regardless of which it is, we can assume it'll be here after Tesla's full autopilot that he promised as well as the Mars colonies.
Full self driving exists, as well as starship. So I take it you mean “imminently”. Like him or loathe him, he executes , which is more than can be said for most.
Earlier quoted context omitted.
Regardless of which it is, we can assume it'll be here after Tesla's full autopilot that he promised as well as the Mars colonies.
Full self driving exists, as well as starship. So I take it you mean “imminently”. Like him or loathe him, he executes , which is more than can be said for most.
[flagged]
Companies have hijacked the open source concept to mean downloadable blob and we follow them as I see in the comments. It’s a real shame.
I remember the NVIDIA Linux kernel binary blob driver discussions from the early-mid 2000s. Who knew we had an open source driver all along...
Earlier quoted context omitted.
Better a downloadable blob than a non-downloadable one
No, it's not, as it means nobody is pushing for actually open models. A truly open model has open code that gathers pre-training data, open pre-training data, open RLHF data, open RLAIF data generated from its open constitution and so on. The binary blob is the last thing I'd want - as a heavy user of LLMs I'm actually more interested in the detail of what all training data is in full, than I am the binary blob.