oofbey

8 ms·

121.
by oofbey 7mo ago
Good point. And diving in I realize my fear was mostly unfounded. Compared to typical background exposure (what we can infer people are getting through other sources by looking at their urine) this is insignificant, except for the very wors
122.
by oofbey 7mo ago
The article is actually IMHO overly conservative. This kind of migration is not a theoretical risk, but well established. BPA is a small molecule, not covalently bound to the plastic. It absolutely goes into the skin. Heat, water, and acid
123.
by oofbey 7mo ago
Thanks for the examples. I was struggling to come up with ideas of how I’d use this.
124.
by oofbey 7mo ago
It truly boggles the mind how bad public schools can be. When you read something like this or see things in person, it’s genuinely difficult to imagine how a system could do this when staffed with well intentioned humans with brains. And ye
125.
by oofbey 7mo ago
Support? This is Google!
126.
by oofbey 7mo ago
Couldn’t be bothered - the Google way. Nobody ever got promoted for doing something easy and useful.
127.
by oofbey 7mo ago
You’re not going to create any future if you insult people trying to offer friendly advice, or think of the talent you rely on to create your vision as “bozo programmers”. I’d wish you good luck, but you have convinced me you don’t deserve
128.
by oofbey 7mo ago
It's not just that crypto lets you cash out faster - it lets you do it with zero notice, accountability, or diligence. Startup exits (IPO or acquisition) often have a big chunk of hype associated with them. But often the hype is backe
129.
by oofbey 7mo ago
Clever comparison, but the key difference is there’s no mechanism for a rug pull for most startups. Unless they reach a huge valuation, the stock is absolutely not liquid. There’s no way to cash out.
130.
by oofbey 7mo ago
Fun fact about laser discs. They are analogue not digital. CD’s store digital information with the presence or absence of pits. Fairly ancient but still fundamentally feels like a very old version of a thumb drive. Laser discs are not digit
131.
by oofbey 7mo ago
I think yours might be extreme. But I think the anonymity here is widely appreciated. And frankly necessarily relies on easy creation of accounts. People share things that they often wouldn’t. And somehow the culture remains mostly civil.
132.
by oofbey 7mo ago
What I find really sad is how slowly this science moves. As I read the article, the “a ha!” breakthrough was in the 1990s. Why did it take them 15-20 years to publish anything??? Coming from the tech industry this is mind boggling. If you
133.
by oofbey 7mo ago
If it wasn’t clear, I don’t have any desire to convince anybody of anything. You don’t believe the future is here yet? Good luck holding on to that position. Not my problem. I was taking time to try to help somebody who sounded genuinely cu
134.
by oofbey 7mo ago
Here’s some practical tips: Start small. Figure out what it (whatever tool you’re using) can do reliably at a quality level you’re comfortable with. Try other tools. There are tons. If it doesn’t get it right with the first prompt, iterate
135.
by oofbey 7mo ago
It’s easy to get AI to write bad code. Turns out you still need coding skills to get AI to write good code. But those who have figured it out can crank out working systems at a shocking pace.
136.
by oofbey 7mo ago
Actually “last year: this”. It was published in early 2025.
137.
by oofbey 7mo ago
First mover disadvantage: you need to invent a way of doing things (ANN) before any decent tools have been built. 10 years later the world is laughing at you for using Java to do vector math.
138.
by oofbey 7mo ago
Not true. The system prompt is clearly different and special. They are definitely trained to differentiate it.
139.
by oofbey 7mo ago
What was the injected title? Why was Claude acting on these messages anyway? This seems to be the key part of the attack and isn’t discussed in the first article.
140.
by oofbey 7mo ago
No doubt! I’ve measured literal 5 minute ping times on airplanes. 300,000ms. Where are the buffering the packets!?
141.
by oofbey 7mo ago
Yet another way Google might decide you’ve violated their ToS and cut you off from your entire digital life without warning or recourse. Sounds handy. But use at your own risk.
142.
by oofbey 7mo ago
In terms of policy this is a truly massive gift to importing companies. They had to pay massive amounts of tax to import goods. Analysis shows most (but not all) of the tariffs were passed on down the line to consumers of the imported goods
143.
by oofbey 7mo ago
The difference in indemnification based on which plan you’re on is super important. Thanks for pointing that out - never would have thought to look.
144.
by oofbey 7mo ago
Thanks for all the detail. I still wouldn't equate "it's going to get a lot worse" with "it's too late". The only way I can interpret the phrase "it's too late" is as giving up - which mea
145.
by oofbey 7mo ago
Good point on speculative decoding techniques. I'd forgotten about them, and they're good. Would love to see some of these get into llama.cpp and friends, but it does require somebody to come up with a distilled draft model. But
146.
by oofbey 7mo ago
Token/s is entirely determined by memory bandwidth. TTFT is compute bound.
147.
by oofbey 7mo ago
You can totally run it locally. If you have 500GB of RAM.
148.
by oofbey 7mo ago
I think this is how it goes: The people who set things on fire weren't directly associated with Greenpeace. But the argument is that Greenpeace's rhetoric incited them to do so. So it becomes a freedom of speech issue. If you t
149.
by oofbey 7mo ago
I think our society would be better off if everybody did less incitement in their political dialogue. It's become all too common for political discourse to become unhinged. So in that sense, I think this precedent is also fine.
150.
by oofbey 7mo ago
Strange take. What does that mean? Give up because we can’t do anything - “it’s already too late”? So burn all the oil because YOLO? Or maybe those people drawing hard lines in the sand were exaggerating to drive urgency and get attention?

More ›