827a

5 ms·

1.
by 827a 10d ago
I don't feel that all fitness trackers are bad, though some use-cases for them are generally bad (such as sleep tracking). Watches are quite good for exercise tracking, and tracking exercise has legitimate and high quality value (e.g.
2.
by 827a 11d ago
There is not a single modern model where, when asked to "Turn the Add To Cart button Blue", would also turn the "Cancel" button next to that add to cart button blue. That's what this site is communicating, literal
3.
by 827a 11d ago
The biggest reason should be that they're just bad. They gather sus data from their undersized sensors, then derive tons of sus insights to overwhelm you. None of it is accurate enough to base major life decisions on, which is also lau
4.
by 827a 17d ago
I don’t know where you picked up the idea that my intention was to be helpful. I’m only mirroring the intentionality behind whoever created this site; they obviously also had no intention of being helpful, as is the case with many discussio
5.
by 827a 17d ago
But its not even good satire, because its totally unrepresentative of my and most others' lived experience. Its similar to making a joke about a calculator misadding two numbers because a stray beam of solar radiation flipped a bit.
6.
by 827a 2mo ago
I'm probably between $50-$200/day depending on the day; we also have effectively unlimited budget, though a lot of that is because Azure gives startups $150,000 in credits for 2 years, which we've wired up to a LiteLLM gatewa
7.
by 827a 2mo ago
Your issue, I believe, is that you seem to believe capabilities are measured along one axis. This is natural to believe because it is representative of how the models have evolved up to this point, and thus it is also what many AGI-pilled p
8.
by 827a 2mo ago
Coding is an extremely verifiable and loopable task, like math (in fact, all of the math that these models has done has been through the lens of Lean, which is itself just coding). I am talking about their capabilities in tasks that are mor
9.
by 827a 2mo ago
I'm very bullish on AI, but I still feel that we pretty much plateaued at Opus 4.6 and everything since then has been in the domain of "extremely verifiable and loopable tasks" (math), benchmarkmaxxing, and harness improvemen
10.
by 827a 2mo ago
Totally untrue. Luna and Sonnet 5 are very comparable: https://artificialanalysis.ai/#intelligence Luna is an extremely strong model.
11.
by 827a 2mo ago
Vera Rubin will be hitting racks very soon, and this is purported to have a 10x improvement in token throughput per megawatt. Of course, old chips don't get replaced with new chips overnight, but I don't think we're anywhere
12.
by 827a 2mo ago
Passkeys are so, so, so bad. One of the worst things our industry invented. The sooner sites start leaving them on the wayside and just go back to TOTP, SMS, and Email codes/links, the better. These work. We solved auth. Its fine.
13.
by 827a 2mo ago
They should definitely fix it, but that's mostly because its an "unnecessary autoplay" so to speak. There's plenty of "necessary autoplays" out there, and AI is going to add more and more every day, because tha
14.
by 827a 2mo ago
Frankly, if you git clone a compromised repository, I'm not sure that a vulnerability of the class "compromised code in that repository will be executed" is all that major a concern. There are plenty of IDEs that will go auto
15.
by 827a 3mo ago
The constitution does not really care about scale, though, and that’s my point. It’s a reason why the legislature should care about Flock, but not why the judicial should.
16.
by 827a 3mo ago
The constitution does not really care about scale, though, and that’s my point. It’s a reason why the legislature should care about Flock, but not why the judicial should.
17.
by 827a 3mo ago
The argument I struggle to get around and would love to hear a counter-argument to: Let's say a local police department hired 175 police officers, each being told "Go stand on this particular intersection with a pad of paper and w
18.
by 827a 3mo ago
Microcenter has them marked at $4500 right now (that's with the 4TB SSD) [1]. I suspect it comes down to what you're using it for; if you're looking for a general purpose computer that's also solid at AI, the AMD machine
19.
by 827a 3mo ago
Framework, weirdly, overcharges considerably for their SSDs. You can currently get a Samsung 990 Pro 2TB on Amazon for $390; Framework charges $625 for the Sandisk 850x 2TB, which has similar performance (and is being sold on Amazon for $53
20.
by 827a 3mo ago
Tbh we'll see what using it looks like, but the reasoning/cost charts do not look promising. It seems like the only useful reasoning level for Sonnet 5 is Low; medium might trade blows at price/performance with Opus, but anyt
21.
by 827a 3mo ago
Why are you comparing xhigh reasoning between Sonnet and Opus? Of course Sonnet xhigh is cheaper than Opus xhigh, but that isn't the point; the point is that at e.g. 80% accuracy on Opus costs ~$0.45 (medium reasoning) whereas on Sonne
22.
by 827a 3mo ago
This seems really, really stupid. Similar to the weird Zig runtime signature thing from a few months ago ago, it was bound to be discovered, quickly, and all the resellers have to do is find a new domain name that (checks notes) doesn'
23.
by 827a 3mo ago
Apple does not sell a 64GB variant of the M4 Mac Mini. IIRC they never have; its always capped out at 48GB. If you were planning on getting an M5 128GB; just get a DGX Spark (~$4500) or a 5090-equipped machine (~$4500) plus a Macbook Air (~
24.
by 827a 3mo ago
Yeah I just mean that if a business came to them and asked for fifty licenses to the $200/mo plan, OpenAI would tell them to kick dirt and basically pay API pricing. Startups should 100% just be telling their employees they can expense
25.
by 827a 3mo ago
Company-wide their margins are trash (probably negative). They need as much inference margin as they can get to afford the massive training runs. It is likely that we'll see GPT-5.6 reduce API pricing to compete against Anthropic, but
26.
by 827a 3mo ago
IMO: If the project leverages Google branding or authority improperly, then it shouldn't be on github and should not be under active development by Google employees; yet it is. If Google is suddenly alright with the way the project lev
27.
by 827a 3mo ago
> Zitron's numbers don't tell us the real cost of generating tokens but, subject to the assumption that the platforms are not subsidizing the token price, that means Anthropic is subsidizing their enterprise customers by up to
28.
by 827a 3mo ago
Intelligence is maybe a few months behind. But cost sadly is further behind. GLM-5.2 has a deceptively high cost during day-to-day usage for e.g. coding because 1) it has to think a ton more than GPT-5.5/Opus-4.8 to get to competitive
29.
by 827a 3mo ago
> if your server is implemented in a sensible way, lol
30.
by 827a 3mo ago
Its a very well known fact that all of our geopolitical adversaries have sophisticated fake American ID markets. This doesn't stop any of the most dangerous adversaries from getting access to systems protected by this technology. DeepS

More ›