dTal

11 ms·

151.
by dTal 5mo ago
USB-C is very far from a perfect connector. The female side still has a fragile plastic tongue that can break. They also reliably wear out with use, both the cables and the socket. We've all seen them fail. Actually all the USB conne
152.
by dTal 5mo ago
Uh this is apocalyptic for computer security, no?
153.
by dTal 5mo ago
Well that's easy enough - don't apply sneaky pricing when there's two people looking.
154.
by dTal 5mo ago
E-ink price tags are not uncommon. Technology to track individual customers through the store based on smartphone RF is already deployed in many supermarkets. Some stores even do scan-as-you-shop, where the customer scans the item at the sh
155.
by dTal 5mo ago
404
156.
by dTal 5mo ago
On the contrary, I think a "simulator of common objections" that reflects all the blind spots and biases of wider society is an extraordinarily valuable tool for exploring and evaluating new ideas. You might find that, in neatly s
157.
by dTal 5mo ago
The word I have heard is "bullshitting". Lies at least orient themselves with regard to the truth, bullshit floats free
158.
by dTal 5mo ago
I love this software for its reliability (as compared to, say, KDE Connect, which I gave up on after years of frustrated use after it became clear that the developers did not believe there was an issue and it would never improve). I do not
159.
by dTal 5mo ago
Bit older than 8 years but even cramming a working GPS reciever into a phone was a huge, nontrivial achievement.
160.
by dTal 5mo ago
>Every number so far has been wrong No it hasn't.
161.
by dTal 5mo ago
>strongly predicting what will happen in 30 years has always been wrong so far No it hasn't, this is climate change denialist nonsense. In fact no less a figure than ExxonMobil correctly predicted the trajectory of global CO2 levels
162.
by dTal 6mo ago
It's a good way to frame the discussion, with the caveat that for some things that subset is 5% of the population, and for other things that subset is 95%. Is there a threshold? Can we define a principle that covers the entire range? I
163.
by dTal 6mo ago
Ooh, car analogy time! It's kinda like saying a car with a 6L engine will always outperform a car with a 2L engine. There are so many different engineering tradeoffs, so many different things to optimize for, so many different metrics
164.
by dTal 6mo ago
Why? No one else was. The discussion was about OpenAI / Anthropic's lack of moat when there are open weights models that are almost as good. You can host them anywhere you like. Pay a US company to do so if you want.
165.
by dTal 6mo ago
It's the same user and they already answered you: "If you read I’m talking about their service only models." But yes this is a non-sequitor. The original question was "What competitive advantage does OpenAI/Anthropi
166.
by dTal 6mo ago
Won't be much "raw material" left before long, if everyone takes that view.
167.
by dTal 6mo ago
The general case is that our own current relative ignorance on the best way to use and adapt pretrained weights is a short-lived anomaly caused by an abundance of funding to train models from scratch, a rapid evolution of training strateg
168.
by dTal 6mo ago
None of that is true, at least in theory. You can trivially change layer size simply by adding extra columns initialized as 0, effectively embedding your smaller network in a larger network. You can add layers in a similar way, and in fact
169.
by dTal 6mo ago
We are in such different universes that I fear that this will not be a productive discussion; to my eyes LLMs are the most obviously socially transformative technology in my lifetime, up there with "internet" and "smartphones
170.
by dTal 6mo ago
>There is no indication that LLMs are a pathway to meaningful and transformative AI. Reality check, they are already astoundingly meaningful and transformative AI. They can converse in natural language, recall any common fact off the top
171.
by dTal 6mo ago
Practically speaking, it isn't secure; no closed app can be. It receives regular compulsory updates (old versions refuse to work) and there's nothing at all stopping Zuck from sneaking in backdoors targeted at you personally.
172.
by dTal 6mo ago
>Altman's death would change nothing about that fundamental calculus. You'd have to kill probably tens of thousands of people to really put a dent in AI development Your analysis seems to assume that people will remain more afr
173.
by dTal 6mo ago
*Aryan - Ayran is Turkish buttermilk :)
174.
by dTal 6mo ago
I keep seeing comments that refer to Iranians as "brown people" - usually to emphasize their perceived "otherness" by the ignorant, as in this case. But Iranians aren't brown, or Arab apart from a small minority, an
175.
by dTal 6mo ago
>Fixed input-to-output mapping is determinism. Prompt instability is not determinism by any definition of this word It really depends on your perspective. In the real world, everything runs on physics, so short of invoking quantum indete
176.
by dTal 6mo ago
Ah, 'twas a mere jest, a sarcastic jab that of all the manifold builds provided, the most useful is missing - doubtless for good and practical reasons. Nevertheless, worth looking at the Vulkan builds. They work on all GPUs!
177.
by dTal 6mo ago
I love this gorgeous and evocative little time waster and come back to it every now and then. Notes: It starts out buttery smooth but over time its performance slows to a crawl. Changing window geometry seems to do some sort of garbage coll
178.
by dTal 6mo ago
The "car-like handling" is still physically accurate - thrusters automatically align your velocity vector to match your view direction. You can think of it as simply an interface - view direction is both a command and a display.
179.
by dTal 6mo ago
Sort of. They are deterministic in the same way that flipping a coin is deterministic - predictable in principle, in practice too chaotic. Yes, you get the same predicted token every time for a given context. But why that token and not a
180.
by dTal 6mo ago
Nobody said anything about Europeans having a "natural right". Bad enough to derail a conversation with irrelevant political nitpicking, unforgiveable to use a strawman to do so. Boo.

More ›