anon373839

5 ms·

61.
by anon373839 2mo ago
I feel like the industry has quietly moved past the Bitter Lesson. In 2023 the story was naive parameter/data scaling and “emergent” intelligence properties. But there wasn’t enough data or compute to keep pushing in that direction, a
62.
by anon373839 2mo ago
But I thought Anthropic was an altruistic organization devoted to the betterment of humanity…
63.
by anon373839 2mo ago
It seems to me that the model struggles to have enough general intelligence, knowledge, or reasoning capacity for arbitrary prompted tool calling. At this size, not surprising. I am VERY interested in seeing how it could perform with some f
64.
by anon373839 2mo ago
Actually I have been quite curious about this. I have a Qwen 3.5 800M model kept in memory just for this type of processing, and I set my small_model setting to use it. But it isn’t receiving any requests and somehow the titles are still ge
65.
by anon373839 2mo ago
> albeit at a lower tok/s rate Not always! I get 80-100 tok/s from Qwen 3.6 35B-A3B on a MacBook Pro thanks to MTP. With long contexts that dips to around 50-60. However, prefill is much slower than API models. So it becomes re
66.
by anon373839 2mo ago
I almost feel like he is just trolling for attention to justify his job. He posted a fake-outrage meltdown tweet about commodification of LLMs being the enemy of America or some nonsense. Paragraphs without a single uppercase character. In
67.
by anon373839 2mo ago
Recurring subscription fees are not “tools” the employee owns, they are employer business expenses. There are statutes requiring employers to fully reimburse business expenses incurred by the employee. Example: Cal. Lab. Code § 2802.
68.
by anon373839 2mo ago
LLMs aren't finished products, though. And since they are basically a distillation of the collective works of all of humanity, it would make a lot of sense for them to wind up as public goods rather than some special advantage held by
69.
by anon373839 2mo ago
I don’t quite understand what the argument is. Government does as a matter of fact operate on the frontier of new technology. (It’s how we got the web.) Why shouldn’t it, exactly? Also, the US has been involved in AI research since the 1940
70.
by anon373839 2mo ago
I haven’t looked at their docs in a little while but they were very precise in saying that the providers follow a zero data retention policy. But Opencode Go proxies the requests, right? I never saw a statement in the legal docs that woul
71.
by anon373839 2mo ago
Commoditizing AI models serves the interests of just about everybody except for a relative handful of people in San Francisco. The more decentralized control of the technology is, the more its benefits can be realized by businesses and indi
72.
by anon373839 2mo ago
Yes. There is a big question mark about whether Anomaly itself (as the OpenCode Go middleman) retains data. The docs were completely silent on this the last time I checked.
73.
by anon373839 2mo ago
Some US states have laws protecting employees’ political speech, so it would be illegal to terminate them and they would be entitled to sue.
74.
by anon373839 2mo ago
If you think the public rejection of AI is weird, wait till you catch a whiff of the marketing! The future described by these AI labs isn’t like previous waves of technological progress. With those, technology displaced some/many occup
75.
by anon373839 2mo ago
A trademark dispute in isolation isn’t related. But as part of a broader course of conduct in which the pursuit of commercial advantage consistently trumps other values, it is. We can disagree on this.
76.
by anon373839 2mo ago
> Isn't the harness basically where the frontier model companies can capture value and create a moat of sorts? They're trying. I see a few main avenues: 1. Fitting the models to their specific harness, so that if you want peak
77.
by anon373839 2mo ago
Let me try putting this in simpler terms: When someone says they are doing something to help you, but it really helps them, you have to choose if you trust what they say. One way to know if you can trust them is to watch how they behave.
78.
by anon373839 2mo ago
Direct and unrehearsed doesn't seem like the right tack for a high-stakes legal dispute between two (allegedly) trillion-dollar companies. Especially since they've not been able to make a coherent point here.
79.
by anon373839 2mo ago
That location was incredible in Nope !
80.
by anon373839 2mo ago
Phones are communication devices. So I would always expect connectivity to be central to their use. But that's a different issue than the "inefficiency" of edge computation and whether powerful edge devices can sell (they obv
81.
by anon373839 2mo ago
> Apple also accuses Tang Tan of trying to get and use their trade secrets. However, Tang has always been clear with the team that we do not want, and must not use, any confidential information from other companies. > Apple’s request
82.
by anon373839 2mo ago
I suspect the economics favor centralized servers, if you only look at the aggregated cost to serve X number of users' tokens. But we could say the same thing about a lot of the computation that iPhones do locally. They could have been
83.
by anon373839 2mo ago
> Another interesting question is why the frontier labs are piling on pure maths The reason is that the original scaling axes (parameters, training tokens, test-time compute) have saturated already, but RLVR (reinforcement learning from
84.
by anon373839 2mo ago
1. This is not “unrelated to AI safety”. The things they seek in the name of AI safety always seem to be aligned with the path that will eliminate competition and give them maximum economic power. Thus, to disentangle what is sincere and wh
85.
by anon373839 2mo ago
I’m sorry, but this do-gooder concept doesn’t jibe with their other actions. For example, swindling their business partners (such as Figma). Filing an unnecessary trademark lawsuit against a customer whose brand predates Anthropic. And f
86.
by anon373839 2mo ago
Not the GP, but I run this model as daily driver too. It runs great on a Macbook Pro 64GB (M3 Max). Token generation speed can be about 100 tokens/sec with multi-token prediction, although it depends on the context. Worst case speed is
87.
by anon373839 2mo ago
Apple Silicon. But: there's no need to use the FP16 version. At 8-bit precision the quality loss is almost imperceptible. That cuts the footprint to 36GB. Which is great for a 64GB Mac, because you have room for plenty of context. 6-bi
88.
by anon373839 3mo ago
Could they even afford 4B dense parameters touching every token? That seems like a LOT of compute compared to generating the classic Google SERP.
89.
by anon373839 3mo ago
Well, my position is that the software that defines/runs the model is the open-source part. The weights are the asset that makes it useful, but I don’t think 40T tokens of pre-training data should be required to call the model itself o
90.
by anon373839 3mo ago
The LLM marketing loop is getting awfully long in the tooth. I saw a meme on Twitter the other day that showed a circular state diagram with something like: > “GPT solved a math problem” -> “Claude solved a math problem” -> “GPT e

More ›