fastball

17 ms·

421.
by fastball 1y ago
Not only is YYYY/MM/DD unambiguous, but it also sorts correctly by date when you perform a naive alphabetical sort.
422.
by fastball 1y ago
The author is almost certainly ljharb.
423.
by fastball 1y ago
Possibly, but I do not think Yudkowsky's opinion of himself has any bearing on whether or not the above article is a good encapsulation of why some people are worried about AGI x-risk (and I think it is).
424.
by fastball 1y ago
I don't think working memory has much at all to do with sentience. The conversation was more about long-term memory, which has not been sufficiently studied in animals (nor am I certain it can be effectively studied at all). Even then
425.
by fastball 1y ago
It’s a new major because they are using it to deprecate other models.
426.
by fastball 1y ago
https://www.lesswrong.com/posts/uMQ3cqWDPHhjtiesc/agi-ruin-a...
427.
by fastball 1y ago
How do you know we have better memory than other animals?
428.
by fastball 1y ago
Top-level async (I assume you mean `await`) is available in browsers. https://caniuse.com/?search=top%20level%20await
429.
by fastball 1y ago
How is that at all relevant? Can we please stop injecting Elon Musk into every single HN conversation?
430.
by fastball 1y ago
Don't know what you're on about. I use Figma to design UI/UX all the time and have found great value in the RTC features. In fact like you say, I was using Sketch before (and tried Framer for a while), but Figma's collab
431.
by fastball 1y ago
Can we stop publishing diet studies that use self-reported diet journals/questionnaires as the method for controlling variables? They are completely unreliable and invalidate anything they touch. In this particular study it is even wor
432.
by fastball 1y ago
> Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting. https://news.ycombinator.com/newsguidelines.ht
433.
by fastball 1y ago
But doesn't this obviate the use of a specific protocol? The protocol itself does very little to help with this problem – only laws with teeth do that.
434.
by fastball 1y ago
Are the use of the word fuck and the aggressiveness separate aspects? And no, I do not know exactly why. In what universe is SpaceX a company that mostly just cares about "increasing shareholder value" and not literally pushing th
435.
by fastball 1y ago
Yeah, the entire tech industry is a monolith which collectively supports one platform or another.
436.
by fastball 1y ago
Yes, basically nothing stays the exact same quality over time. But if you can get better quality and worse quality that kinda obviates the OP's point.
437.
by fastball 1y ago
- I didn't say a single entity, I said "owners". In the case of X there are many owners, Elon Musk with the most power. Nobody who is not an owner controls the company, which is what you would need to demonstrate to invalid
438.
by fastball 1y ago
I'm seeing those numbers, but as the low end (without storage) for solar and at the high end for nuclear... so not a reasonable comparison. Not sure where you're looking, but from here your numbers are way off base. To be more con
439.
by fastball 1y ago
Where are you getting a LCOE diff of 5x? The latest Lazard's is 2x. Transmission costs will require more than "wiggle room" if you are sending power from some cornfield in middle America to Seattle. Also a big question in my
440.
by fastball 1y ago
The fundamental disagreement here is (in my opinion) on what needs to be encouraged. If what you say is true and renewables are cheaper anyway, then also as you say capitalism should in out and that's what we will use in the near term
441.
by fastball 1y ago
More expensive in what way? "Cost" is what everyone quotes about why nuclear isn't great, but isn't the whole idea behind shelving fossil fuels and switching to alternatives due to downsides that are secondary to cost? T
442.
by fastball 1y ago
Or we can go full nuclear.
443.
by fastball 1y ago
Yes, owners of companies obviously control them. That doesn't make the CEO a figurehead. There isn't a company in existence where the (hired) CEO has more power than the person who owns the company .
444.
by fastball 1y ago
Sure.
445.
by fastball 1y ago
It is actually not that hard to make something water-tight if you have no intention of opening it again. > I dont think nuclear power is a solution. It can be step in an energy transition strategy, but no solution. Do you mean nuclear fi
446.
by fastball 1y ago
Good point. I suppose the answer here then is to not allow the Supabase MCP any exfiltration pathways.
447.
by fastball 1y ago
It's only a security hole if you give access to users though, right? If you are the one using the Supabase MCP, how is it any different than any other root access to a DB?
448.
by fastball 1y ago
Database providers do just throw a postgres prompt at developers though, right? And that is what Supabase is – an infra provider.
449.
by fastball 1y ago
This whole "don't train on my data" thing is so silly. Do you know how these models were created? By training them on code. Very selfish / tragedy of the commons for you to want to use tools that were trained on the code
450.
by fastball 1y ago
ICYMI, DeepMind also has a Gemini model that is diffusion-based[1]. I've tested it a bit and while (like with this model) the speed is indeed impressive, the quality of responses was much worse than other Gemini models in my testing. [

More ›