Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
qlte
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
121.
▲
by
qlte
1mo ago
Isn't this basically the phenomenon known as low-end disruption/upmarket migration? [1] Which is often viewed as an unhealthy sign for a company that recognizes it can't keep up with efficiency of a new market entrant but unw
122.
▲
by
qlte
1mo ago
Nothing about watermarking would require padding the response length with pseudo-intelligible Claudese. Regular filler would work fine. Also, it would probably provide higher entropy to write normal human-sounding English instead of reusing
123.
▲
by
qlte
1mo ago
Okay, finally a plausible explanation for why Anthropic has stubbornly appeared to be self-sabotaging their model with the overly broad yet somehow also overly specific biology filter. The fear of Trump Admin retaliation excuse made superfi
124.
▲
by
qlte
1mo ago
It's pretty wild how "reasoning" models now generate like 10 thousand hidden chain of thought tokens in response to a "increase opacity of the logo by 20%" prompt before writing the actual message and yet they stil
125.
▲
by
qlte
1mo ago
I've had a Boox Page for about two years now as a Kindle Oasis replacement (current model is Boox Go 7 B&W). I get a bit more than a week of battery life typically which is good enough for me. And most valuable is using the same th
126.
▲
by
qlte
1mo ago
If they're jailbroken.
127.
▲
by
qlte
1mo ago
Boox Go 7 (B&W version) or Boox Page (discontinued) if you're willing to go Android. Both have the Oasis form factor with two physical buttons. Battery life is more like 1.5 weeks but hasn't been an issue for me at all. For me
128.
▲
by
qlte
1mo ago
https://archive.ph/MISxW
129.
▲
These Generals Fought for Israel. Now They See 'Jewish Terrorism' as the Threat
(nytimes.com)
21 points
by
qlte
1mo ago
|
1 comments
130.
▲
by
qlte
1mo ago
That's totally true, I have no issues with a changing opinion over time. The problem is expressing both opinions with 100% certainty and not updating priors to consider that your new position could also be completely wrong.
131.
▲
by
qlte
1mo ago
I agree but don't think that's the best example, although not directly alignment related LW loves that kind of ideation of fantastical sci-fi scenarios. Better would be the very evident negatives of non-ASI AI the world is already
132.
▲
by
qlte
1mo ago
I mean if doesn't help that Eliezer Yudkowsky spent the first decade of his public life zealously spreading the gospel of the AI singularity as the solution to all mankind's problems. And then later did a 180 degree pivot to AI si
133.
▲
by
qlte
1mo ago
> Same deal with states with "stand your ground laws". It's astonishingly easy to not get shot: don't invade someone's home. Florida road rage incident leaves 2 girls shot after fathers' stand-your-
134.
▲
by
qlte
1mo ago
> Just disconnecting from Wifi often doesn't work because the TV will connect to insecure wifi to upload data Is there any evidence of this actually happening? It's oft stated yet I've never seen any proof or speci
135.
▲
by
qlte
1mo ago
Slow-walked, conveniently, until the admin has finished doing whatever the original lawsuit claimed was illegal, and then the government argues it is impossible to undo after the fact.
136.
▲
by
qlte
1mo ago
No, this isn't just an overused rhetorical device like all politicians do, those examples are the flimsy legal argument the Trump Admin uses in court, frequently as justification to break the actual letter of the law.
137.
▲
by
qlte
1mo ago
Sure, I can imagine it. But I had a Windows Mobile phone and definitely didn't need to do this, there was a Phone button on the start screen.
138.
▲
by
qlte
1mo ago
Scaling means something different in this context IMO. It's not whether your DB can go from serving 100 to 10000 users, it's whether your B2B app can serve 100 to 10000 different real world customers without starting to fall apart
139.
▲
by
qlte
1mo ago
> then they will never work and SpaceX will fail Or, Elon pulls his classic narrative sleight of hand for like the 9th time and announces SpaceXesla is NOT a car, battery, solar, self-driving, robot, space launch, space datacen
140.
▲
by
qlte
1mo ago
The user has to first specifically pick the device from the list and grant the website access. If a user is confused by a permissions prompt and has no idea what is going on, the default path is to reject the permission.
141.
▲
by
qlte
2mo ago
Pretty revealing how you had to hand wave away the entirety of "Liberation Day" tariff policy (literally the topic of this thread) that Biden did not do in office to do the perennial low-info voter "uniparty" observati
142.
▲
by
qlte
2mo ago
I believe we are several generations past peak-RLHF at this point. Now it's much more RLVR (Reinforcement Learning with Verifiable Rewards), with a goal/evaluator loop. Which, conveniently, fits neatly into the benchmaxxing arms r
143.
▲
by
qlte
2mo ago
> but devices like the Steam Deck prove that mobile linux isnt as much of a problem re: userspace as you claim. So.... by relying on yet another company with multiples of $10B in revenue who spent a lot of time and money to build
144.
▲
by
qlte
2mo ago
AOSP security model relies on an extremely restrictive SELinux config out-of-the-box (that importantly also doesn't impede normal phone/app usage), with the kernel hardened about as much as any cutting edge Linux distro could offe
145.
▲
by
qlte
2mo ago
Or rather, linked to completely wrong negative health effects that were just “common knowledge” thus requiring no actual evidence like: People with light skin should always wear a hat that fully covers the head/face in tropical/e
146.
▲
by
qlte
2mo ago
Don't really get the incredulity considering he needs to justify the tens of billions already spent on AI infrastructure by, you know, training a model or two ...
147.
▲
by
qlte
2mo ago
According to the article, the “application” (which I imagine was also broken down into smaller pieces as they went back and forth to regulators during the review process) was 12,000 pages in total. The 2 million number is for “supporting” m
148.
▲
by
qlte
2mo ago
Shouldn’t have accepted the sloth in the first place. They may look cute and docile, but also have a panic reflex to grab onto the nearest tree-like thing they can feel with their claws, including their would-be attacker. Or if they just fe
149.
▲
by
qlte
2mo ago
No, that's the "preview" version (gemini-3.1-pro-preview) aka the beta/early version before the official release of 3.1 Pro.
150.
▲
by
qlte
2mo ago
Seems less like "bad"/broken defaults and more defaults tuned to the max for benchmarks. All the positive PR from "Opus 4.6 level" online buzz is well worth the minor annoyance from taking a half hour to solve a sim
More ›