Certhas

6 ms·

31.
by Certhas 2mo ago
Posts like this truly boggle the mind... The US is in a class of its own when it comes to health spending. The second highest OECD country (per GDP) is Germany which is 5 points lower than the US (and BTW these figures include public and pr
32.
by Certhas 2mo ago
The US spends more than 16% of GDP on health. The UK around 11%. Shrinking US health expenditure by 3% of GDP while roughly maintaining health outcomes is eminently, obviously doable.
33.
by Certhas 2mo ago
I don't doubt that ISO models are not optimal, but the annual cost of power outages to US consumers is above 100 billion dollars last time I looked already. Electric system reliability is a place where you have to be somewhat conservat
34.
by Certhas 2mo ago
Elon Musk has actively endorsed the right wing extremist party in Germany. The one that often has trouble running it's events as businesses try to avoid serving them. And which has prominent members that the secret service considers de
35.
by Certhas 2mo ago
The design and intellectual heritage of Julia and Mojo couldn't be more different though. Mojo is a Swift/Rust style language from industry veterans, Julia is a Lisp built to make academics write faster code.
36.
by Certhas 2mo ago
The Python tie in was always a sales pitch. The technical pitch was "a language built around MLIR taking all the lessons learned by Rust and Swift on board, with an eye towards Python interop". There were MLIR intrinsics and exten
37.
by Certhas 2mo ago
I am using many many many things that I don't understand. Cars, public transport, etc... I review and test the end product, not every tiny step along the way. If the LLM uses some command line tools I have never heard of to create a mo
38.
by Certhas 2mo ago
I agree in principle. The tech community naturally has a tendency to reach for tech solutions to societal problems. But without jurisdiction, laws are meaningless. If you create a law that says you have to honour robots.txt files, what do y
39.
by Certhas 2mo ago
Evaluating LLMs is incredibly difficult. They are categorically different from any other system we have intuition about. That said, I read the question I am replying to as a rhetorical one. If it was meant as a genuine question, curious abo
40.
by Certhas 2mo ago
How do humans?
41.
by Certhas 2mo ago
So for every work produced by AI have ten separate agents review it thoroughly.
42.
by Certhas 2mo ago
In research we are still seeing massive jumps. Subjects that LLMs were completely useless for half a year ago are now definitely in scope. And there are benchmarks that cleanly separate the SOTA models: https://epoch.ai/Mirr
43.
by Certhas 2mo ago
Roughly Anthropic owners benefit from the content, unless the scans are made available somehow.
44.
by Certhas 2mo ago
I think the example makes it pretty clear that this is exactly what is meant. I was thrown by the term interactivity at first as well. But I think the trichonomy is sound, and the Erlang Vs Python Vs Rust buckets are quite meaningful, thoug
45.
by Certhas 3mo ago
Trade barriers are a different thing than immigration though. I believe there is much more evidence that trade displaced traditional jobs than that immigration led to an oversupply of labour.
46.
by Certhas 3mo ago
This is not evidence. Just naive theory, probably LLM generated.
47.
by Certhas 3mo ago
As far as I am aware this is a racist/xenophobic narrative with little to no evidence. Automatisation is responsible for job losses, not immigration. If you have studies/evidence saying otherwise please share. I agree that too man
48.
by Certhas 3mo ago
I can be amazed and terrified at the same time.
49.
by Certhas 3mo ago
Not true. These seats are being phased out. Enterprise is now token only.
50.
by Certhas 3mo ago
And I know many people that don't. That have a 20 or 100 dollar subscription for very bursty workflows with months where they barely use tokens. Not every subscriber is a full time SWE. In fact most professional SWEs will be on enterpr
51.
by Certhas 3mo ago
This is hotly debated and completely unclear. Let's say Anthropics Opus models cost the same to serve as GLM 5.2. GLM 5.2 is 4.4$/MTok while Opus is 5.6 times more expensive. Assume that GLM 5.2 is served at essentially zero margi
52.
by Certhas 3mo ago
Please post an example of a response that shows such masking.
53.
by Certhas 3mo ago
* * *
54.
by Certhas 3mo ago
I read the article you linked: https://www.chinatalk.media/p/how-to-buy-cheap-claude-tokens... I know that Anthropic claims that "large-scale coordinated "distillation attacks"" exist. They are not
55.
by Certhas 3mo ago
This topic really is a testament to people's willingness to opine on things they have absolutely no clue about. A first year undergraduate can completely check this counterexample in ten minutes. The original post even linked Wolfram a
56.
by Certhas 3mo ago
I mean, China is the leader in PV panels, battery tech, telecommunication hardware, etc... already...
57.
by Certhas 3mo ago
First off, I don't doubt that China engages in large-scale industrial espionage, or at least used to. Nowadays they have plenty of talented and highly educated engineering talent of their own. Second, I now actually read the article. I
58.
by Certhas 3mo ago
TOS violations are not espionage. Everybody who links up Claude to OpenCode is violating the TOS. So from the largest industrial espionage in history we have left "They paid for the accounts but violated the TOS". And then you ran
59.
by Certhas 3mo ago
You haven't explained how this is illegal or any more immoral than scraping the web for training data. As you said yourself: They are buying the product. Then they are using it for their own purposes. That's more than Anthropic&#x
60.
by Certhas 3mo ago
Can I interact with and discover any federated instance without having to know it exists? My experience with Bluesky Vs Mastodon really showed that the friction of federation in the latter can really kill the experience for me. I think we n

More ›