oh_no

7 ms·

1.
by oh_no 8d ago
It still works, bots can solve it but it probably increases the cost of that web call by 10x or 100x for that bot, so it won't bother. Had a recent bad experience with removing recaptcha.
2.
by oh_no 8d ago
I get this pretty frequently on windows Firefox after switching to it, note this is my work computer, Firefox works fine at home on more open network
3.
by oh_no 13d ago
Very nice to see that this is even more token efficient than Sol, when Fable 5.1 is less so than the already bloated token budget of Fable 5.
4.
by oh_no 13d ago
that's all openai models but i'm very happy openai continues to focus on efficiency rather than reasoningtokenmaxxing
5.
by oh_no 14d ago
the last thing nvidia would want is to make ai look more legally risky to say nothing of their massive investments direct and indirect in openai
6.
by oh_no 15d ago
with Fable 5.1 increasing token use pretty dramatically I'm again impressed that OpenAI seems like the only lab to be driving token use down. The ExploitBench Internal Port chart showing token usage is crazy impressive
7.
by oh_no 19d ago
you dont like giving openai money but you're fine with spacex? not criticizing, just trying to understand.
8.
by oh_no 19d ago
cursor adds $0.25/M to your token bill for 3rd party services. sounds small but it's insanely big on cached inputs, which are an insanely high % of use
9.
by oh_no 23d ago
1. I think LLMs will end up pretty dramatically shrinking MTTR, a lot of that will be tooling to proactively resolve problems as soon as they start, but a lot of it is that agents are very good at finding and fixing problems by correlating
10.
by oh_no 23d ago
very easy to lose money on subscription, very easy to make money on api pricing
11.
by oh_no 23d ago
what? enterprise LLM API contracts pay listed model rates. how could they lock you into pricing on models that they've not developed yet? enterprise SaaS LLM calls do lock in rates but don't lock in models for similar reasons to t
12.
by oh_no 29d ago
yes and no, anthropic and openai are losing money on people who max out their sub, but openai has a lot more room to play with with much cheaper models to serve (by all signs we have from actual api/task pricing)
13.
by oh_no 1mo ago
5.6 Luna costs far less and benchmarks far better, have you compared for this task?
14.
by oh_no 1mo ago
buddy, they're on enterprise plans paying per token
15.
by oh_no 1mo ago
i'm a little confused by this, should most piles of company documentation look pretty similar? what are we getting by tuning at the org level? what if you have a bunch of teams or apps that have different documentation patterns? how mu
16.
by oh_no 1mo ago
few days old but want to flag that there are zero apples:apples comparisons on this press release. set aside the benchmark vs older models they're testing their model with access to their internal legal research data vs open internet s
17.
by oh_no 2mo ago
where are you seeing cheap Kimi? pricing I've seen is the same across the board (presumably due to licensing terms) and is in the Terra range.
18.
by oh_no 2mo ago
I pretty strongly disagree about comparing this to Kimi and GLM, 5.2 was a big price hike for Chinese models, and Kimi K3 was a big price hike to that. K3 was within spitting distance of OpenAI pricing (more expensive than short context Ter
19.
by oh_no 2mo ago
reading some of the comments i was expecting something really high end and polished, but yeah, this is something you could with a headset 1/10th the cost
20.
by oh_no 2mo ago
based on pricing I think it's safe to say they're different. why would they charge half price when fable has been very popular?
21.
by oh_no 2mo ago
seeing a jump this big is not a great sign for the continuing value of a benchmark
22.
by oh_no 2mo ago
I have also thought about this a lot but have less developed views on fiction (short version: I don't currently want to read an LLM-written novel, but I expect that will change in the future and there will be some I enjoy) But I want t
23.
by oh_no 2mo ago
GLM 5.2 and Kimi 3 both had huge API pricing jumps (GLM most expensive chinese model, by a lot, then then Kimi 3 a lot higher than that). the cost advantage is rapidly decaying. the oft-cited cost per task makes Kimi look good vs Anthropic
24.
by oh_no 2mo ago
it has kimi 2.5, which isn't to say this will show up but who knows
25.
by oh_no 2mo ago
i don't think so, i think it's 50% what work people are doing, 50% vibes. my experience with 5.5 is i like it more and get better results than 4.8/fable. which isn't to say i think it's a strictly better model, just
26.
by oh_no 2mo ago
top of the line SSDs now eclipse DDR3 throughput, but DDR3 should retain a large edge in latency of orders of magnitude absolutely no idea how useful any of that would be and what kind of latency degradation going through whatever adapter w
27.
by oh_no 3mo ago
Foundation didn't create the dataset, just the framework for volunteers to do the work.
28.
by oh_no 3mo ago
so i think you're a bit off. it's s/g but g is legit accounts who want to buy the steam machine. we could say it's 5000 scalper accounts, and 50000000 gamer accounts. but it's not 5000/50000000, it's like
29.
by oh_no 4mo ago
except google does respect robots.txt so you do have a choice?
30.
by oh_no 5mo ago
interesting that GPT Image-2 managed to 2-shot this with thinking turned on, I didn't save a copy and it disappeared from my window but I first got a failure very similar to the one in the article, but it saw the issue and said it was

More ›