xracy

16 ms·

271.
by xracy 2y ago
What do we count as "using AI"? How much AI do you have to use for work to be considered "using AI"? I used AI once for work 1 year ago, when it was popular to try, does that count? I remember when Alexa was considered &
272.
by xracy 2y ago
"What can be asserted without evidence can also be dismissed without evidence" By this razor, I feel like I can dismiss your opinion. Can you elaborate why you think he doesn't know enough about technology for you to pay at
273.
by xracy 2y ago
That's not the start of an argument. It's the start of them explaining their point.
274.
by xracy 2y ago
IDK, I think it's kinda important for LLMs to get the simple stuff correct in order to justify looking into the rest of the hallucinations. Like, if you can't tell me "what day is it today?" (actual failed prompt I have
275.
by xracy 2y ago
> I see no reason that a language model couldn't do iterative science in the same way humans have been with the resources available to it (APIs) using the composition of current technologies. I think this belies a common refrain fr
276.
by xracy 2y ago
Realistically a bus is 2-3 times the size of a car. I think you need like 5 people to justify it. And 6 before you're actively addressing traffic. Most bus capacity is around 10. Which is 4 cars saved. Also, do you care about the state
277.
by xracy 2y ago
Traffic congestion increases nonlinearly[1]. i.e. the 10th car on a road has a smaller effect than the 100th car on the road. Accordingly, the cost of increasing a bus' size and decreasing its speed pales in comparison to the value der
278.
by xracy 2y ago
Yeah! The thing that's gonna fix this systemic issue is more personal responsibility! Just stop using youtube everyone! OR pay for youtube premium!
279.
by xracy 2y ago
I love me some linux distros, and I would sure love to play all of the steam games I like on a real Linux os... And if that's actually improved, I'll sure consider it. And if it's not, I don't really have a choice other
280.
by xracy 2y ago
Feels like what AI should be used for... "filtering out the noise" rather than creating it.
281.
by xracy 2y ago
I mean, I guess that's a possibility, but it's also impossible to prove, and the graphs of "what might've been" are probably more likely to be incorrect than correct.
282.
by xracy 2y ago
I mean, fusion energy is still 10 years out. Just like it was a decade ago.
283.
by xracy 2y ago
That would've made a better case if they hadn't tried to hire ScarJo for this. This looks like: 1. They wanted ScarJo's voice 2. She said 'No' 3. They then proceeded to ask themselves the question you're asking
284.
by xracy 2y ago
I would be very concerned about what this is saying. I don't think it means these women experience less harassment. I suspect it means they have a higher tolerance for it from experiencing more of it. To me, this says we have a long wa
285.
by xracy 2y ago
I feel like there are probably diminishing returns on picking your best metric... Like, does it matter if the number is 5x as deadly or 2.5x as deadly when you consider that not adjusted for population we have a ton of traffic deaths in
286.
by xracy 2y ago
Actually this is not ridiculous.
287.
by xracy 2y ago
Depends on the courtroom.
288.
by xracy 2y ago
I know this is a popular take right now, but it's actually just wrong. The 1st amendment absolutely protects your right to protest in the US. But practically you do have to be in the right courtroom for that to be meaningful.
289.
by xracy 2y ago
Literally just gainsaying someone on the internet about their insurance. And your gainsay is hedged without actually providing them with some resource they could use to inform them? You're really making a contribution with this comment
290.
by xracy 2y ago
> if you're pissing off the person you're trying to get your message to, you're still shooting yourself in the foot. "Squeaky Wheels Get the Grease" And if they're inconveniencing you, suddenly the thing you
291.
by xracy 2y ago
I wonder how long it will be before we realize we can't have nice things. In the meantime, this seems cool!
292.
by xracy 3y ago
No, but based on Uvalde it wouldn't actually be helpful either.
293.
by xracy 3y ago
> If HR/PR lied about defacing property it would open them up to a defamation lawsuit. The entire purpose of HR is to protect the company from lawsuits which is why they don't usually make up fake reasons for firing people. The
294.
by xracy 3y ago
Your mistake is assuming what that means rather than inferring it from what's actually described as the actions. It's not your fault, but it's wrong to assume that the article writer is unbiased in this situation given that i
295.
by xracy 3y ago
> “obviously, the people mentioned … aren’t [female] at all”, but “men performing their idea of femaleness”. the bracketed word there I replaced is [women]. The remarkable thing about this sentence, is how wrong it is with the word it ch
296.
by xracy 3y ago
I would actually argue it's strictly worse, because we have gotten terribly effective weapons in the last century. No one in the ME was killing each other as effectively as they are now. That's a way things are demonstrably worse
297.
by xracy 3y ago
This take appears to have been "too hot for HN to handle" For everyone downvoting, I'd be interested in the proportion of articles that discuss race as a percentage of total articles. If the number is more than the number of
298.
by xracy 3y ago
How does one address racist policies without discussing race? That's a knot I can't tie myself into.
299.
by xracy 3y ago
On one of my other comment threads someone told me how the world is "so much better then any point in history" :). I'm not sure I believe that. I think in general being able to have these conversations means that we can progr
300.
by xracy 3y ago
I think I have some bad news for you. The media is incredibly biased with their own agenda. I think the only way you can truly account for that is to make up for it with multiple viewpoints. Maybe LLMs can write truly "unbiased" n

More ›