Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
AustinDev
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
by
AustinDev
5d ago
gemini 3.8 flash?
2.
▲
by
AustinDev
6d ago
Hmmm, I've never actually checked that. Just had codex run the numbers excluding the boilerplate in some of the documents output is on average ~13% shorter than input, I only checked the ones that were all of the same type (proposals)
3.
▲
by
AustinDev
6d ago
Here's how I use it to write. I dump an unedited stream of consciousness into the prompt. Sometimes it's just me having a conversation with myself thinking through what I actually want to say. Then, I have it ask me to clarify any
4.
▲
by
AustinDev
8d ago
Pure slop, completely unadulterated from what I can tell.
5.
▲
by
AustinDev
10d ago
It doesn't really matter it's all the same thing for the point of the discussion. All investigations into these 'hacks' are focusing too much on the model and not enough on what the humans did wrong. The model doesn
6.
▲
by
AustinDev
11d ago
Digging back into the HF report. It looks like the initial prompt told Claude that it was in a simulated environment. However, there is also evidence from the traces that the bots figured out that they were not in the sandbox but kept using
7.
▲
by
AustinDev
11d ago
>If the model is prepared to break the rules when it knows it's being observed why should we trust it when it's not being observed. That's fair enough.
8.
▲
by
AustinDev
11d ago
Funny enough if the model thought it was on the real internet it likely would not have done any of these 'hack' events. The model believing it was in a sandbox is why it behaved the way it did (against its normal alignment rules)
9.
▲
by
AustinDev
11d ago
Irregular purports to be a cybersecurity firm and their founders have ties to Anthropic and Effective Altruism. CEO, CTO and other founders sit on various boards for EA organizations.
10.
▲
by
AustinDev
11d ago
If your service doesn't let my agent transact with you then I won't use your service. I'm sure Amazon will sort this out in some way it's hard to replace their infrastructure but it could happen if they fail to.
11.
▲
by
AustinDev
15d ago
The five eyes have four other eyes. But, from where I'm sitting the primary eye has only gotten stronger in the last 2 years. See some work we did in late 00's to early 2010's called Nexus 7 and extrapolate to modern capabili
12.
▲
by
AustinDev
16d ago
Yeah, back when this stuff was pretty new in 2024 I found a jailbreak and sent some bio/chem weapon instructions it generated to an organic chem PhD friend. They told me not only were the procedures wrong but that there were several st
13.
▲
by
AustinDev
16d ago
If the five eyes intelligence services aren't monitoring all biodefense researchers and bio chemists around the world all of the time they've probably failed in their jobs.
14.
▲
by
AustinDev
16d ago
How probable do you think it would be that a layman following ChatGPT's bio weapon instructions would just blow themselves up or poison themselves? It's not like positive pressure environments and suits are very common let alone a
15.
▲
by
AustinDev
16d ago
It was probably just me trying to figure out if I could put one type of draino down the drain within 30 minutes of using a different type. Sorry y'all.
16.
▲
by
AustinDev
17d ago
Precisely, all I hear is a bunch of cheap talk about something gravely serious, if true. I think it's much more likely that the dread these Anthropic employees are experiencing is reckoning with the fact that they may actually lose the
17.
▲
by
AustinDev
17d ago
If they actually believed this they'd be buying a truckload of fertilizer and diving it to the nearest chip fab. I don't buy it.
18.
▲
by
AustinDev
18d ago
Cannon-armed Vessels were the pinnacle of military technology at the time the Bill of Rights was authored.
19.
▲
by
AustinDev
19d ago
I know you're an idealogical zealot but, for anyone else reading. I do find it interesting that the only laws I can find limiting the arming of private vessels were with respect to arming them and then sending them off to fight in fore
20.
▲
by
AustinDev
19d ago
>"the Second Amendment predates machine guns" That statement is false. Repeating firearms like the Puckle gun predate the bill of rights by ~75 years. More importantly, the Founders were hardly unfamiliar with privately owned m
21.
▲
by
AustinDev
19d ago
Remove the antenna.
22.
▲
by
AustinDev
21d ago
No company will ever value your privacy or security more than or equal to how much you value them. This is why you gotta keep an unencrypted bitcoin private key in your password manager. I'll know pretty quickly (within ~ 10 minutes or
23.
▲
by
AustinDev
26d ago
My bad, I was thinking hormonal.
24.
▲
by
AustinDev
26d ago
> abortion to the margins along with birth control even when HIV became an obvious pandemic. How are abortion and birth control at all related to HIV? I don't understand the connection.
25.
▲
by
AustinDev
27d ago
Put the sulfur back in the maritime fuel.
26.
▲
by
AustinDev
1mo ago
I had it read all of the emails find which accounts were associated with the old email and then go switch them to the new email. I only had to manually intervene for 2 accounts out of 36ish.
27.
▲
by
AustinDev
1mo ago
>I was using a @gmail.com address. Bad mistake on my part. So actually tracking what I used the address with, that was my real pain. This is thankfully mostly solved now by AI. I recently had a do a migration that I had been putting off
28.
▲
by
AustinDev
1mo ago
Alternatively, don't use GH for critical infrastructure they are no longer reliable, simple as.
29.
▲
by
AustinDev
1mo ago
Humanoid robots are obviously more than marketing. The entirety of human civilization is human shaped. Making robots that are human shaped is easier and more efficient than redesigning and rebuilding everything that exists.
30.
▲
by
AustinDev
2mo ago
It would appear that their altruism isn't very effective.
More ›