Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
skybrian
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
16 ms
·
331.
▲
by
skybrian
3mo ago
This blog is kind of an interesting hybrid: > Every article published on SmarterArticles is authored and editorially controlled by Tim Green. Artificial intelligence tools are used within a structured and supervised workflow as research
332.
▲
by
skybrian
3mo ago
Not just you and the article shows changes between Opus 4.6 and 4.7 that seem related.
333.
▲
by
skybrian
3mo ago
Yes, but it might be hard to do better. Labels are one-word summaries for axes that probably don’t perfectly correspond to any English word. It’s a common hazard when converting research results to natural language. The words you use for ve
334.
▲
by
skybrian
3mo ago
No. Google has them too.
335.
▲
by
skybrian
3mo ago
I’m assuming the company has policies about what changes they can support. If that’s not true then it’s more of a hosting service.
336.
▲
by
skybrian
3mo ago
That's why I assume it goes through support. A support person checks if the prompt is reasonable (and maybe rewrites it).
337.
▲
by
skybrian
3mo ago
My guess is that SaaS will be around, but it's going to look really different. Say there's a company that sells you a subscription to an issue tracker. At first, it looks just like any other web-based issue tracker. But, although
338.
▲
by
skybrian
3mo ago
> I’ve struggled for years to imagine this, ever since scaling started for real in 2020, and I failed to get productivity out of chatbot-tuned LLMs, with their creatively-stunted endlessly repetitive prose. Instead, while lagging behind
339.
▲
by
skybrian
3mo ago
Things look different if you’re comparing an American tech company to an American government.
340.
▲
by
skybrian
3mo ago
Maybe the bottleneck will be the people who have to think of new tasks for AI? That is, markets do saturate; customer demand has limits. After it peaks, what will be left for AI to work on?
341.
▲
by
skybrian
3mo ago
These effect systems seem nice enough when all your I/O is going through a narrow interface, but I'm wondering what happens when you have a very large API surface, like everything available to a web page running in a browser?
342.
▲
by
skybrian
3mo ago
While single and working at a startup back in the dot-com days, I shared an apartment in the Richmond district with two other people. It didn't seem all that unusual. I assume people still do that? Seems like it would be okay temporari
343.
▲
by
skybrian
3mo ago
Huh, I've never seen that.
344.
▲
by
skybrian
3mo ago
Arguments based on incentives can prove anything. We could just as easily say they have incentives to provide tools to both sides and let them fight it out. It would be like Gmail automatically whitelisting email from other Gmail accounts o
345.
▲
by
skybrian
3mo ago
That too, but they need competitors to switch to.
346.
▲
by
skybrian
3mo ago
It looks like they have the right idea but their workspaces are based on GitHub worktrees rather than separate VM’s.
347.
▲
by
skybrian
3mo ago
Yeah but I ‘d rather run it remotely and not run agents on my laptop at all.
348.
▲
by
skybrian
3mo ago
Yes, but it’s up to their competitors to build competing services.
349.
▲
by
skybrian
3mo ago
It’s less weird if you think there’s a difference between good bots and bad bots. They can provide services for good bots to use while helping people keep out the bad ones. If a bot is simulating mouse movement but doing it badly then that’
350.
▲
by
skybrian
3mo ago
If the agent is running on your machine, it will suspend when you put your laptop asleep. I prefer using a remote Linux VM to let the coding agent keep working. I’m quite happy with exe.dev for this. My laptop is asleep upstairs but I have
351.
▲
by
skybrian
3mo ago
Relentlessly? Where did he post about this and how often?
352.
▲
by
skybrian
3mo ago
It will be front-page news if there are storms later this year. For now, ocean temperature readings far away are effectively a distant early warning of future weather. (Depending on your perspective about “distant” and “early.”)
353.
▲
by
skybrian
3mo ago
Porting TypeScript to Go in 7.0 doesn’t seem like bad engineering. Neither does this port, given that that Bun started out in a somewhat buggy state.
354.
▲
by
skybrian
3mo ago
When a mathematician reads a hundred-year-old math paper, they are reproducing in their head the reasoning of someone who died long ago. That is, reasoning can be written down and replicated. If that works, I think it's fair to say tha
355.
▲
by
skybrian
3mo ago
It’s often very good at finding posts that I might theoretically want to read, except that I never want to read them on Facebook, because it would get in the way of seeing posts from friends.
356.
▲
by
skybrian
3mo ago
That paper argues that an LLM “lacks the mechanism for Abduction,” which is not the same thing as a claim that “creativity cannot be automated.” They propose a different kind of AI: > The emergence of physically consistent World Models o
357.
▲
by
skybrian
3mo ago
Looks like the Allen Family Philanthropies has given away about $500 million so far and has about $100 million in “active” grants. I guess they will be giving away a lot more? https://www.allenphilanthropies.org/who-we-are#r
358.
▲
by
skybrian
3mo ago
I wonder what the Allen estate will do with the money?
359.
▲
by
skybrian
3mo ago
Some YIMBY laws have passed, but it takes time to build things. For example, it looks like the 25-story Marina Safeway can't be stopped, though they're still trying. Berkeley has quite a few new buildings, though.
360.
▲
by
skybrian
3mo ago
It's abrupt because I didn't copy the entire article, just quoted a few paragraphs. I changed it to an ellipsis.
More ›