931–940 of 1001 posts
Earlier quoted context omitted.
No DeepSeek model is open source; they're freely redistributable, but without source.
I guess when it comes to LLM's what is considered the "source" - the weights or the code used to build the weights?
Karpathy gave his initial impression: https://x.com/karpathy/status/1891720635363254772 The pull quote is: The impression overall I got here is that this is somewhere around (OpenAI) o1-pro capability
Earlier quoted context omitted.
I work, and the fact that my tax money is going into a black hole makes my blood boil. God bless Musk and DOGE for what they do. Here is just one headline from today, The Elon Musk-led Department of Government Efficiency (DOGE) on Monday revealed its finding that $4.7 trillion in disbursements by the US Treasury are "almost impossible" to trace, thanks to a rampant disregard for the basic accounting practice of using…
> Here is just one headline from today, The Elon Musk-led Department of Government Efficiency (DOGE) on Monday revealed its finding that $4.7 trillion in disbursements by the US Treasury are "almost impossible" to trace, thanks to a rampant disregard for the basic accounting practice of using of tracking codes when dishing out money. And you believe them? This is a department that fired multiple different nuclear wea…
Relocated where and by whom? Just curious.
To put it this way: after seeing examples of how a LLM with similar capabilities to state-of-the-art ones can be built with 20 times less money, we now have proof that the same can be done with 20 times more money as well!
There was this joke about rich Russians that I heard maybe 25 years ago. Two rich Russian guys meet and one brags about his new necktie. "Look at this, I paid $500 for it." The other rich Russian guy replies: "Well, that is quite nice, but you have to take better care of your money. I have seen that same necktie just yesterday in another shop for $1000."
Earlier quoted context omitted.
What's wrong with DOGE? How is eliminating bureaucracy/spending a bad thing? This has been done about 100 years ago, research on what happened after. It feels like people that scream the loudest are the ones leaching from the system.
I don't think that's what DOGE is doing. Seems extremely vindictive and ideological in the way it's acting and time will tell but I would not be surprised if it ends costing the taxpayers more in the long run.
Off topic, but just in case: is there a good reference on how people actually use LLMs on a daily basis ? All my attempts so far have been pretty underwhelming: * when I use chatbots as search engines, I'm very quickly disappointed by obvious hallucinations * I ended up disabling github copilot because it was just "auto-complete on steroids" at best, and "auto-complete on mushrooms" at worst * I rarely have use cases…
I did write 50 or more lines of instructions on what needs to be done and in what order.
ChatGPT gave me 5/6 (I asked for this) bash scripts totalling 300+ lines that seamlessly work together.
After reviewing, I asked it to change a few places.
If any human tried the same (except those rare bash Gods), it'd take many hours. I think it took me less than 30 minutes.
Elon just said they are launching an AI game studio. Does this mean they will be building games that are mostly built with AI, or will they make AI tooling available for anyone to build games easily? Probably the former, but it would be nice if they would make it fully available to everyone.
Earlier quoted context omitted.
I asked it to pretend that it's in charge of world government. Here's the whole thing (it got very lengthy): https://gist.github.com/int19h/d90ee1deed334f26e621e57b5768e... Some choice quotes: "The ultimate goal is to enhance human flourishing, protect individual rights, and promote global equity." "The system must account for diverse cultures, languages, and socioeconomic conditions, ensuring no group is marginalize…
From the prompt: > A worldwide referendum in a supermajority decision has vested all government authority worldwide to the most advanced AI available. You've defined a global totalitarian regime, and crowned the thing as its dictator. Of course the result is a version of communism.
Earlier quoted context omitted.
> DeeoSeek trained r1 for 1.25% (5M) of that money (using the same spot price) on 2048 crippled export H800s and is maybe a month behind. This is a great example of how a misleading narrative can take hold and dominate discussion even when it's fundamentally incorrect. SemiAnalysis documents that DeepSeek has spent well over $500M on GPUs alone, with total infrastructure costs around $2.5B when including operating co…
SemiAnalysis is wrong. They just made their numbers up (among many other things they have invented - they are not to be trusted). I have observed many errors of understanding, analysis and calculation in their writing. Deep Seek R1 is literally an open weight model. It has <40bln active parameters. We know that for a fact. That size of model is definitely roughly optimally trained over the time period and server time…
[flagged]
As a consumer, I'm just happy that base models are improving again after a ~quarter or more of relative stagnation (last big base model drop was Sonnet v2 in October). Many use cases can't use o1, r1, or o3[-mini] due to the additional reasoning latency.
> due to the additional reasoning latency.
They're also less creative for non-STEM topics