xpe

50 ms·

721.
by xpe 1y ago
I get your point. Still, there are other ways to respond, even if they are difficult. One is to be curious and take what you can from it. Another is to think about people poorer and less fortunate than yourself who might look at you dimly f
722.
by xpe 1y ago
Can you summarize how you see the Web3/blockchain aspect and its centrality for the role?
723.
by xpe 1y ago
I wouldn’t phrase it as a matter of “validity”. I would phrase it as a question of transparency. Putting a model out in public without clearly explaining how it works doesn’t meet my bar for a proper scientific exchange of knowledge. Perhap
724.
by xpe 2y ago
> Companies are revenue maximizers, period. This explanatory model explains a lot of what companies do but not all. It is a useful first approximation for many firms. Still, the conceit of modeling an organization as a rational individua
725.
by xpe 2y ago
Yes, OpenAI demonstrated one way not-for-profits can be commandeered. Altman appears to be quite astute at gaining power. Every organizational design and structure has the potential to be subverted. Like cybersecurity, there are many tradeo
726.
by xpe 2y ago
I agree, but I might phrase it a little bit differently. I recommend thinking about corporate stances as actions and interests, not moral intentions. Don’t expect a corporation to do things for moral reasons. Trust them only to the extent t
727.
by xpe 2y ago
>> These benchmark numbers cannot be real for a 7b model > LLM benchmarks are mostly bullshit right now. Wait a few years until the hype cycle returns to sanity. This could mean a lot of things. Can you be a bit more specific? It&#
728.
by xpe 2y ago
The README says "RL" without specifying what kind of RL is used. Researchers: I know you are busy, and I know good writing takes time, but please don't skip this kind of detail.
729.
by xpe 2y ago
To answer directly: Ask a question. Watch the “thinking” process. Estimate the likelihood that all of the generated text is in the training data. Do you disagree with my claim? Or perhaps you were hoping for a very rigorous set of experimen
730.
by xpe 2y ago
> Theories are developed by doing the work and LLMs do not do the work. They ingest the output of work. This is often the case but does not _have_ to be so. LLMs can use chain of thought to “talk out loud” and “do the work”. It can use s
731.
by xpe 2y ago
This kind of debate isn't that unusual, is it? To figure out what calculation to use, one has to made tradeoffs about desirable properties such as resilience to outliers, being bias-free, and/or having the best predictive performa
732.
by xpe 2y ago
> I truly do not understand why this administration continually gets the benefit of doubt from the HN audience. I'm guessing you're not understanding me. Regarding the strength and validity of the FBI's criminal complaint,
733.
by xpe 2y ago
>> I'm not a lawyer, but the criminal complaint does appear to be, more or less, within the realm of normal. > Criminal complaints against sitting judges for actions they took in their courtroom are not at all "normal&quo
734.
by xpe 2y ago
I think the claim underneath the comment above is that immigration laws have been not been enforced or unevenly been enforced. On principle, advocating for a consistent application of the law seems sensible. This goes along with advocatin
735.
by xpe 2y ago
Generally, I share these concerns. At the same time, this story is very new. In any case, looking at the primary sources is important. See https://www.courtlistener.com/docket/69943125/united-states-... . I'm
736.
by xpe 2y ago
See also section D that says "Judge DUGAN escorts Flores-Ruiz through a “jury door” to avoid his arrest" in https://storage.courtlistener.com/recap/gov.uscourts.wied.11... labeled "Case 2:25-mj-00397-SCD
737.
by xpe 2y ago
If you want a primary source, I recommend reading https://storage.courtlistener.com/recap/gov.uscourts.wied.11... In particular, part D: "Judge DUGAN escorts Flores-Ruiz through a “jury door” to avoid his arrest.&
738.
by xpe 2y ago
Want to try a conversational reset? I'll start. My disagreement with the claim "AIs are overeager junior developers at best" largely has to do with both understanding what is happening under the hood and well as personal expe
739.
by xpe 2y ago
We have very different ideas of what "literal" means. You _interpreted_ what I wrote as "just Google it". I didn't say those words verbatim _nor_ do I mean that. Use a search engine if you want to find some high-qua
740.
by xpe 2y ago
I’m thinking you might be a kind of person that requires very direct feedback. Your flagged comment was unkind and unhelpful. Your follow-up response seems to suggest that you were justified in being rude? You also mischaracterize my commen
741.
by xpe 2y ago
Or uppercase would have cleared it up.
742.
by xpe 2y ago
> Gemini is ... way cheaper. Yep. Here are the API pricing numbers for Gemini vs Claude. All per 1M tokens. 1. Gemini 2.5: in: $0.15; out: $0.60 non-thinking or $3.50 thinking 2. Claude 3.7: in: $3.00; out: $15 [1] https://ai.
743.
by xpe 2y ago
In terms of short-term outlay, perhaps. But don't forget to factor in the long-term benefits of having a human team involved.
744.
by xpe 2y ago
You can set spend limits https://docs.anthropic.com/en/api/rate-limits
745.
by xpe 2y ago
Good point. I largely work in Zed -- looks like it had semantic search for a while but is working on a redesign https://github.com/zed-industries/zed/issues/9564
746.
by xpe 2y ago
by fe the poster means FE (front-end)
747.
by xpe 2y ago
> So, AIs are overeager junior developers at best, and not the magical programmer replacements they are advertised as. This may be a quick quip or a rant. But the things we say have a way of reinforcing how we think. So I suggest refinin
748.
by xpe 2y ago
> GitHub copilot follows your context perfectly. I don't have to tell it anything about files. I tried this initially and it just screwed up the results. Just to make sure we're on the same page. There are two things in play. F
749.
by xpe 2y ago
Don't forget to add the CCP to the party. (DeepSeek has ties to the CCP.)
750.
by xpe 2y ago
> and likely influenced by the fact that the company is Chinese Let's rephrase this. It is not simply that DeepSeek is a Chinese company. It is because of its links to the CCP [1] [2] and the CCP's cyber operations. [1]: https

More ›