brookst

10 ms·

211.
by brookst 2mo ago
Are you really advocating for assuming things with no evidence, by presenting no evidence for why one should do so? That’s not especially rigorous thinking.
212.
by brookst 2mo ago
It was never that dramatic, and it’s declining day by day. It’s a panic over a real but small problem. Order of magnitude more water is lost from wasted irrigation (e.g. during rain, of fallow fields, sprayed into windy air, etc) than data
213.
by brookst 2mo ago
Citation please? Bartz v Anthropic seems pretty clear, see also Authors Guild v Google and RIAA v Diamond Multimedia.
214.
by brookst 2mo ago
Wait, so the library of congress is suddenly responsible for probably petabytes a day of incoming scans? To what end? Do they have to index it and make it available? Do they have to check the accuracy and integrity of the scans? How does th
215.
by brookst 2mo ago
No, that’s the hyperbolic reaction clickbait wants from you. Not all rare books are valuable. Someone’s self-published junk sitting in the garage is NOT analogous to the library of Alexandria. Many, most, maybe all of these “rare” books are
216.
by brookst 2mo ago
The dissonance when a cause you support is loudly represented by disingenuous types. At some point they’ll hit on data center water consumption as yet another reason to support the cause.
217.
by brookst 2mo ago
Because human commits ar necessarily quality?
218.
by brookst 2mo ago
I’m getting a lot more done. Hobby projects that languished for years are coming along great, at quality and depth I could never have found time for before.
219.
by brookst 2mo ago
Yes, exactly. As compared to 70 years ago, when cars came with a manual on how to adjust the spark gap, and buyers were expected to do so. It turns out most people don't want to be under the hoods of their cars, so cars are optimized f
220.
by brookst 2mo ago
All of these terms were picked by individuals, years ago, while reaching for metaphors that made sense to them personally. None of these "agent" / "thinking" / "reasoning" terms were dreamed up in boa
221.
by brookst 2mo ago
But how is that any different than people being misled by real humans saying words that reflect real thinking, but which are actually dead wrong? The fallacy here is "thinking == correct", not "tokens == thinking"
222.
by brookst 2mo ago
It really is. I know I've gone from tens a month to thousands a month. They have to be projecting >100B/month in the next year or two.
223.
by brookst 2mo ago
Partly fair, though I’d argue I think less of people who treat their car badly to express their own bad mood. But we also interact with cars differently than LLMs. The way I’m typing a note to you, here, is much more similar to interaction
224.
by brookst 2mo ago
But the semantic meaning can be agent-specific. For instance, my Claude.md files have things like “never present multiple questions in walls of text, use `AskUserQuestion` with clear tradeoffs described”
225.
by brookst 2mo ago
While they’ve certainly fucked up goodwill with their actions, it’s also true that the tech enthusiast community has always been like teenagers who reject their favorite band when it gets popular. Oh, I used Claude before they got popular…
226.
by brookst 2mo ago
While I dearly love Kafka, I am not sure it’s correct to look at him as a documentarian of realistic social behaviors.
227.
by brookst 2mo ago
How is it different from relying on the internet, when most people don’t know what a routing table is, let alone how to manage one? Or from relying on word processors that the average user could never write themselves? Civilization has alwa
228.
by brookst 2mo ago
While I agree it’s theoretically possible, do you believe this capability exists in the LLMs we use today?
229.
by brookst 2mo ago
Why would those things make you interact with it differently? I’ve always believed that being kind and conscientious is beneficial to ourselves. I’m certainly not interested in using LLMs to practice skipping those behaviors myself.
230.
by brookst 2mo ago
So, a requirements document for updates to a legacy system?
231.
by brookst 2mo ago
For Linux? Assuring stakeholders that they can both control and observe development and how it works. For a charity? Assuring donors that funds are being managed appropriately. For Anthropic? No benefit at all. Turns out "transparency&
232.
by brookst 2mo ago
I think you believe there's an equal-outcome promise in the law that really just doesn't exist. Not everyone gets caught. The deterrence is the likelihood of getting caught times the penalty. And yeah, the guy with a bright red, l
233.
by brookst 2mo ago
Important by what measure?
234.
by brookst 2mo ago
M5 AVP could probably render thousands of them in realtime
235.
by brookst 2mo ago
I mean what if there’s a mountain at that height? Will they just fly into it?
236.
by brookst 2mo ago
You think their name means releasing competitive details would be good for them? I’ve got sone bad news about Federal Express.
237.
by brookst 2mo ago
Doesn’t any enforcement action “single out” a wrongdoer? People pulled over for speeding are singled out (there are other speeders), people busted insider trading are singled out (there are other inside traders), etc? I honestly don’t under
238.
by brookst 2mo ago
In fully supportive of prosecuting wrongdoing in the food industry. But it seems strange to use belief in widespread wrongdoing in the food industry to oppose punishing wrongdoing in the tech industry. Is it just nihilism? All industries ar
239.
by brookst 2mo ago
Law is about intent. It is not code, and it’s a mistake to think of law like code where you need to unit test and cover all edge cases. In law, it is sufficient to show an intent to take actions that foreseeably caused harm. Specific laws,
240.
by brookst 2mo ago
Why would they? What the upside, for them?

More ›