News for hackers. Stuff that matters.

throwpoaster

Re:StackOverflow now gets fewer than 1000 questions a month on Thursday October 29, 2026 @07:40AM (#50287444)
And all get closed as duplicates, right?
A null transform is a transform.

Engage seriously with the project and OP will deliver.

Re:Where are YC founders now? OpenAI and Anthropic, mostly on Thursday July 16, 2026 @12:34PM (#48936787)
> the fact that the entire economy is betting everything on it

This is untrue. Spend is about 2% of GDP.

Could be a great buying opportunity depending on your theory of investment.
Re:What will be left for us to work on? on Tuesday July 14, 2026 @04:39AM (#48903813)
Chop wood. Carry water.
With or without AC?
Re:Rocketlab acquires Iridium on Tuesday June 30, 2026 @04:22AM (#48729893)
Looks awesome!
The steel man is that you can’t peer review your way to breakthroughs that change consensus, because peer review relies on consensus, so peer review has to be made subordinate to accountable decision makers.
Re:Mechanical Watch on Tuesday June 16, 2026 @12:13PM (#48557547)
NNNOOO000ooo...…
Re:Mechanical Watch on Tuesday June 16, 2026 @10:29AM (#48555890)
This (re-)triggered a huge (dormant?) mechanical watch special interest for me when it was first posted.

Neurospiceys tread carefully.

<3

Nice, I once built a ray tracer without using an SQL database.
Maybe we changed the rules so that our capital markets can protect our champions from large Chinese buy pressure at lower post-IPO valuations. Plus, these companies have probably had to demonstrate earnings stability (yes, unaudited) in order for the rules to be waived.

Is the text of the waiver and its reasoning anywhere? I guess I'll read tfa.

Re:MCP Is Dead on Saturday May 30, 2026 @10:40PM (#48342563)
No, the technical opinion. MCP is an interesting integration pattern. Calling it "more APIs" is like calling software "just some code". It's true, not interesting.
Re:MCP Is Dead on Friday May 29, 2026 @08:54PM (#48331250)
I suggest you implement an MCP server before adopting this as a firm technical opinion.
Re:Flipper One – we need your help on Thursday May 21, 2026 @10:07AM (#48222829)
Protip: asks need to be simple. This is cool, but long: "congratulations, or sorry, but I'm not reading that".

If they had a "preorder" button at the top I would give them money and be done with it.

Re:Shunning AI is the human choice on Thursday May 21, 2026 @10:03AM (#48222770)
Pass. Hate is never good.
Re:There Is No 'Hard Problem of Consciousness' on Monday May 18, 2026 @09:54AM (#48179945)
“Post hoc, ergo propter hoc.”
Re:Agents for financial services and insurance on Tuesday May 05, 2026 @03:13PM (#48027146)
Because the l on your keyboard is broken?
Re:Talking to 35 Strangers at the Gym on Monday May 04, 2026 @11:15AM (#48009756)
Autism is a hell of a drug.
Interesting that Baumol's Cost Disease might have missed that non-productive wage increases might actually swamp productive wages.

If we assume that services jobs will be, primarily, the jobs that survive automation, the cost disease might also swamp the Coasean Singularity.

I value correctness in my argumentation, not popularity. What part of it don’t you understand?
Yes, although as the Koch Brothers point out in their book: you have to play by the rules that exist, not the rules you want.

If you read, eg, Buffet, he makes the point that a manager donating to a political cause, whether the Heritage Foundation or, God forbid, something as far right as the SPLC, makes that donation with money that otherwise accrues to the shareholders. The manager therefore creates an agency problem, where he might pursue his own interests at the expense of the owners.

If they are aligned, the manager can retain the earnings and create a dividend for the owners, such that they can then make the donation directly. If they are not aligned with the owners, they are redistributing wealth.

I am not surprised that the Left advocates for backdoor wealth redistribution, but I would prefer they be honest about it.

Legal agreements are explicit.
What kind of model "reproduce"s things later for profit that is not generative?
That's not what the Friedman Doctrine is, technically. It is that management should obey moral, ethical, and legal frameworks in the operation of the business for the benefit of its investors; and specifically NOT take actions which are outside of that narrow scope.
Booo, nihilism is dumb.
One important researcher fabricated data that directed too much funding to the amyloid hypothesis.

Next question.

Re:Work with the Garage Door Up on Thursday April 23, 2026 @09:40AM (#47875662)
Seems related to the explore/exploit problem, where the standard answer is related to the answer to the Secretary Problem[0], with the important caveat that it depends on whether "passing" on an opportunity legitimately makes it unavailable in future.

But another good answer is to open the door and trust the audience. The people who show up to the garage practice are perhaps not people who show up to buy tickets.

Adopting a scarcity mindset, generally, is a bad idea.

[0]:https://en.wikipedia.org/wiki/Secretary_problem

Older »