Live data from Hacker News

Google CEO says more than a quarter of the company's new code is created by AI

businessinsider.com

931–940 of 1001 posts

Re: Google CEO says more than a quarter of the company's new code is created by AI

#931

I work for Google, and I just got done with my work day. I was just writing I guess what you'd call "AI generated code." But the code completion engine is basically just good at finishing the lines I'm writing. If I'm writing "function getAc..." it's smart enough to complete to "function getActionHandler()", and maybe suggest the correct arguments and a decent jsdoc comment. So basically, it's a helpful productivity…

I also work at google (until last Friday). Agree with what you said. My thoughts are 1. This quote is clearly meant to exaggerate reality, and they are likely including things like fully automated CL/PR's which have been around for a decade as "AI generated". 2. I stated before that if a team of 10 is equally as productive as a team of 8 utilizing things like copilot, it's fair to say "AI replaced 2 engineers", in my…

Google has to sell its AI some how. The problem is that businesses will see this and want to fire head count because they go, "Well I guess AI can do it for freeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee!". Nope no way is it writing code freely.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#933
post #112
post #94

Earlier quoted context omitted.

You can come up with measures for it and then watch them, that’s for sure.

when metric becomes the target it ceases to be a good metric. when discovered how it works developers will type the first character immediately after opening the log. edit: typo

That was my point, veiled in an attempt to be cute.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#935

Earlier quoted context omitted.

>> Even if that's been happening, I don't think it would be politically savvy to admit it. When I was working in RPA (robotic process automation) about 7 years ago, we were explicitly told not to say "You can reduce your team size by having use develop an automation that handles what they're doing!" Even back then we were told to talk about how RPA (and by proxy AI) empowers your team to focus on the really important…

> Even back then we were told to talk about how RPA (and by proxy AI) empowers your team to focus on the really important things. As a non-politically savy person ;-) I have a feeling that this is a similarly dangerous message, since what prevents many teams to focus on really important things is often far too long meetings with managers and similar "important" stakeholders.

The reason you don't lead with headcount reduction is two-fold.

1. Almost every business has growing workload. That means reassigning good employees and not hiring new headcount, not firing existing headcount. Unipurpose, low-value offshore teams are the only ones who get cut (e.g. doing "{this} for every one of {these}" work).

2. Most operational automation is impossible to build well without deep process expertise from the SME currently performing it. If you fire that person immediately after automating their task, what do you think the next SME tells you, when you need their help?

Successfully scaling operational automation programs therefore rely on additional headcount avoidance (aka improving their volume:employee ratio) and value measurement (FTE-equivalent time savings) to justify/measure.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#936
post #750

Earlier quoted context omitted.

The public might though be concerned that if they are being replaced, many in other positions at other companies will soon be replaced as well.

That’s not how the mind works. People cheered when Elon fired 80% of the Twitter staff. No one cares when people with high paying jobs suffer.

I think twitter is an outlier because people hated the employees already for various reasons.

For example they thought that twitter had a bloated workforce because of videos like this (https://www.youtube.com/watch?v=buF4hB5_rFs).

And a lot of people heavily disagreed with how they handled moderation. You can take things like the hunter Biden laptop suppression or in the funny category you had the getting banned for saying learn to code (https://reason.com/2019/03/11/learn-to-code-twitter-harassme...).

Take random company without controversies and you will find less vitriol about them getting fired.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#937
post #465

Earlier quoted context omitted.

> Most programming is trivial That's a bold statement, and incorrect, in my opinion. At a junior level software development can be about churning out trivial code in a previously defined box. I don't think its fair to call that 'most programming'.

Think of all the menial stuff you must perform regardless of experience level. E.g. you change the return type of a function and now you have to unpack the results slightly differently. Traditional automated tools fail at this. But if you show some examples to Cursor, it quickly catches on to the pattern and start autocompleting semi-automatically (semi because you still have to put the cursor to the right place but…

Don't misunderstand. I am not making an assertion that GenAI tools for development are useless.

I am just pointing out that the thread parent started his logical climb at a step one that is incorrect: 'Most programming is trivial'.

Given that they got it wrong on step one, how good do you thing step ten is?

Re: Google CEO says more than a quarter of the company's new code is created by AI

#938
post #725
post #636

Earlier quoted context omitted.

I'm working on a CRM with a flexible data model, and ChatGPT has written most of the code. I don't use the IDE integrations because I find them too "low level" - I work with GPT more in a sort of "pair programming" session: I give it high level, focused tasks with bits of low level detail if necessary; I paste code back and forth; and I let it develop new features or do refactorings. This workflow is not perfect but…

> I paste code back and forth There is this tool Aider. Takes your prompt, adds code files (sometimes not all of your code files but files it figures relevant) and prepares one long prompt, sends it to an LLM, receives the response, and makes a git commit based on the response. If you rather review git commits, it can save you the back-and-forth copy-pasting. https://aider.chat/

Note that the default mode will automatically change and commit the code, which I found counter-intuitive. I prefer using the architect mode, where it first tells you what it is going to do, so you can iterate on it before making changes.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#939
post #596

Earlier quoted context omitted.

To err is human. To err at scale is AI.

I fear that we'll see a lot of humans err at scale next Tuesday. Global warming is another example of human error at scale.

>next Tuesday.

USA (s)election, I guess.

Re: Google CEO says more than a quarter of the company's new code is created by AI

#940
post #815

Earlier quoted context omitted.

To err at scale isn't unique to AI. We don't say "no software, it can err at scale".

CEOs embracing the marginal gains of LLMs by dumping billions into it are certainly great examples of humans erring at scale.

yep, nano mega.
Post reply on HN