oivey

26 ms·

211.
by oivey 3y ago
No, it isn’t a massive conspiracy. A handful of executives decided to score some political points by publicizing a demonstrably farcical narrative around organized retail theft.
212.
by oivey 3y ago
Their businesses are failing because they’re mismanaged and online retailers are putting them out of business. Blaming a made up retail theft boogeyman distracts from their incompetence and hopefully keeps stock prices and golden parachutes
213.
by oivey 3y ago
Not once in the article does it mention tax cuts. It mentions tax abatements, which are very much not the same thing. Any abatement of this size is going to be inherently political. I would be surprised if someone at GM used this web form t
214.
by oivey 3y ago
It isn’t a tax cut. It’s a tax subsidy unevenly distributed to one company. Extremely not free market.
215.
by oivey 3y ago
Arguably no one’s goal is to glue pieces of C together. That’s too abstract. Their goal is to do something like write a performant web server, maybe easily and quickly. Python’s approach is one way to do that, but there are other ways that
216.
by oivey 3y ago
I’m not sure this is true. Nvidia ships hardware today that is specialized for transformers, and presumably AMD needs something similar on their side. An arms race of specific hardware like that is probably not to their advantage. It makes
217.
by oivey 3y ago
Why are you ok with that? Adding a software governor to the motors just to drag more money out of something you already bought is crazy.
218.
by oivey 3y ago
People get mad at grade inflation because GPA is used to make job/internship selections. A high GPA in Yale’s biology program mostly indicates that you were able to get into Yale, not that you’re a good biologist. That flies in the fac
219.
by oivey 3y ago
> Or, put another way: you can undo effects of a lens No, you most definitely cannot. The roots of computational photography are in things like deblurring, which in general do not have a nice solution in any practical case (like non-zero
220.
by oivey 3y ago
This is the definition of anthropomorphism: “Anthropomorphism is the attribution of human traits, emotions, or intentions to non-human entities.” By definition the point you’re making doesn’t make sense.
221.
by oivey 3y ago
You would need to get a certificate of occupancy, yes. You would also need permits for some unfinished work, too. You brought up the exterior needs to be finished. Do you expect to move in with just the frame? It seems fair to me to have re
222.
by oivey 3y ago
You can do that generally in the US. You’re not allowed to do things like move into a house without running water, toilets, etc.
223.
by oivey 3y ago
I think we’re in agreement mostly. I was trying to make the point that the model needs to contain some practical assumptions at the level we’re discussing, or it breaks down. Both binary integer addition and binary address look ups are O(L
224.
by oivey 3y ago
Fair enough that perhaps you can allow M to be a function of N. It feels like this sort of model might run into other issues, though. If you treat a memory look up as an operation on an infinitely long bit string that therefore takes non-co
225.
by oivey 3y ago
I think you’re confusing some variables here. If you say mergesort is O(N log N), N is the number of elements in the collection. The size of the address space is orthogonal to the limit being expressed. If a memory look up is O(log(M)) wher
226.
by oivey 3y ago
It’s going to cost millions of dollars to set the fees to zero and increase the sticker price? Why would anything new need to be built?
227.
by oivey 3y ago
That database already exists and many new cars have it anyway. My mid range Honda from 2018 has it, for example. Clearly, Honda and Acura wouldn’t need to develop such a thing, and I’m guessing neither would virtually any other manufacturer
228.
by oivey 3y ago
The median new car price in the US was $46k. The amortized cost of such a component would be a rounding error. Not only that, as another commenter and TFA point out, cars already have software limiters installed to prevent engine damage. Se
229.
by oivey 3y ago
Almost certainly adding a speed limiter to a modern car would cost effectively $0. This is a dumb regulation, but “regulation always makes products more expensive” trope is silly, too.
230.
by oivey 3y ago
> Well, if allll someone needs is to detect giant missiles roaring into the edge of space on a pillar of fire, then we should be ultra-suspicious if they start building a system which goes beyond that into, say, following around the hot-
231.
by oivey 3y ago
If core talent leaves I’ll agree. This was a business guy.
232.
by oivey 3y ago
It’s a performance critical optimization that has downsides. It’s a completely fair criticism. It’s like someone claimed a hammer is a universal construction tool, and then someone else pointed out hammers aren’t great for screwing in scre
233.
by oivey 3y ago
Right, so it isn’t intuitively a hard task. It’s hard in the framework of ChatGPT’s design. That design is majorly powerful, but isn’t a great one for this basic task.
234.
by oivey 3y ago
The non-tech media and public have literally no idea who Sam Altman is and realistically whatever happens here isn’t going to change whether people like ChatGPT or not.
235.
by oivey 3y ago
Seemingly there is this consensus of board members around a senior executive. It just isn’t the CEO.
236.
by oivey 3y ago
I’m sure he’s not a dumb guy, just disposable relative to OpenAI’s engineering team. I doubt he’s a Jobs-like, indispensable visionary, either.
237.
by oivey 3y ago
Wasn’t he more of a business guy while Ilya was the engineer? I really doubt a random VC guy is going to really know much about the specific, crucial details the engineering team knows.
238.
by oivey 3y ago
What is Sam Altman going to do with a GPU? Find some engineers to use them I guess?
239.
by oivey 3y ago
Altman’s job was to be a go between for the business and engineering sides of the house. If the chief engineer who was driving the company wasn’t going to communicate with him anymore, then he wouldn’t serve much of a purpose.
240.
by oivey 3y ago
I think this is longer off than you might expect. LLMs work because the “answer” (and the prompt) is fuzzy and inexact. Proving an exact answer is a whole different and significantly more difficult problem, and it’s not clear the LLM approa

More ›