News for hackers. Stuff that matters.
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @02:21AM (#46702186)
> A principal engineer at Google posted on Twitter that Claude Code did in an hour what the team couldn’t do in a year.

I’ll bring the tar if you bring the feathers.

That sounds hyperbolic but how can someone say something so outrageoulsy false.

Re:Ask HN: Do you have any evidence that agentic coding works? by thornewolf on Wednesday January 21, 2026 @02:52AM (#46702453)
as someone who worked at the company, i understood the meaning behind the tweet without the additional clarification. i think she assumed too much shared context when making the tweet
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @02:59AM (#46702499)
A principal engineer at Google made a public post on the World Wide Web and assumed some shared Google/Claude-context. Do you hear yourself?
Re:Ask HN: Do you have any evidence that agentic coding works? by tokioyoyo on Wednesday January 21, 2026 @06:02AM (#46703886)
Working in a large scale org gets you accustomed to general problems in decision making that aren’t that obvious. Like I totally understood what she means and in my head nodded with “yeah that tracks”.
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @09:30AM (#46706215)
Maybe it helps them sleep at night.
Re:Ask HN: Do you have any evidence that agentic coding works? by grayhatter on Wednesday January 21, 2026 @11:11AM (#46707671)
People make mistakes, it's not that deep. The correct incentive to encourage is admitting, and understand and forgiving when necessary because you don't want to encourage people to hide mistakes out of shame. That only makes things worse.

Especially considering forgetting the delta between yours and someone else's shared context is extremely common. And the least egregious mistake you can make when writing an untargeted promo post.

Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @12:09PM (#46708441)
My bad. I will be more mindful tomorrow when someone at a big tech company yet again make-a-mistake in the same direction of AI hyping. Maybe with a later addendum. Like journalists that write about a Fatal Storm In Houston and you read down to the eighth paragraph and it turns out the fatality were among pigeons.

> when writing an untargeted promo post.

lol.

Re:Ask HN: Do you have any evidence that agentic coding works? by grayhatter on Wednesday January 21, 2026 @01:24PM (#46709458)
> My bad. I will be more mindful tomorrow when someone at a big tech company yet again make-a-mistake in the same direction of AI hyping.

Are you mad at them for playing the game, or mad that that's the game they have to play to advance at their company?

> Like journalists that write about a Fatal Storm In Houston and you read down to the eighth paragraph and it turns out the fatality were among pigeons.

I don't know; I guess I hold people who post on twitter so they can self-promo, or who have attention because they work at $company, to a slightly different standard than I would hold a journalist writing a news article?

Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @04:53PM (#46712108)
> I don't know; I guess I hold people who post on twitter so they can self-promo, or who have attention because they work at $company, to a slightly different standard than I would hold a journalist writing a news article?

I know. One of them has a higher salary.

Re:Ask HN: Do you have any evidence that agentic coding works? by xboxnolifes on Thursday January 22, 2026 @01:11PM (#46722992)
Do you think people who work at Google are perfect?
Re:Ask HN: Do you have any evidence that agentic coding works? by JimDabell on Wednesday January 21, 2026 @06:34AM (#46704199)
Who are you referring to here? If you follow the link, you will see that the Google engineer did not say that.
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Wednesday January 21, 2026 @08:25AM (#46705388)
I am quoting the person that I responded to. Which linked to this: https://xcancel.com/rakyll/status/2007659740126761033#m

> I’m not joking and this isn’t funny. We have been trying to build distributed agent orchestrators at Google since last year. There are various options, not everyone is aligned... I gave Cloud Code a description of the problem, it generated what we built last year in an hour.

So I see one error. GP said “couldn’t do”. The engineer really said “matched”.

Re:Ask HN: Do you have any evidence that agentic coding works? by woooooo on Wednesday January 21, 2026 @11:50AM (#46708180)
The key words in the quote are "not everyone is aligned". It's not about execution ability.
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Thursday January 22, 2026 @02:08PM (#46723753)
The key words are “one hour”. So what if there is some mealy-mouthed preamble?
Re:Ask HN: Do you have any evidence that agentic coding works? by woooooo on Friday January 23, 2026 @04:10AM (#46730259)
At big tech companies "were still finding alignment" is code for leadership not being able to make decisions and unblock execution.
Re:Ask HN: Do you have any evidence that agentic coding works? by keybored on Friday January 23, 2026 @06:25AM (#46731207)
Dearest leadership not unblocking execution. What's that got to do with what we’re complaining about? Did Claude cure cancer (this is hyperbole) in one hour or not?
Re:Ask HN: Do you have any evidence that agentic coding works? by woooooo on Friday January 23, 2026 @11:10AM (#46734183)
The point of the story to me was, with a clear idea of what they want, the person got a year's worth of Big Tech done in a few hours with Claude. Could have been a couple days with a tight team, either way, the problem wasn't the coding ability or typing speed.