HackerTrans
TopNewTrendsCommentsPastAskShowJobs

yarekt

no profile record

comments

yarekt
·21 giorni fa·discuss
I get it, and it draws parallels with debugging and incident response in software dev. Your best tool is good data, if you have more data, you’re able to make good judgement in short time. If AI helps you collect good data about your condition, then presenting that to a clinician is certainly better than “it hurts, idk” because there’s only so much they can do at the time
yarekt
·2 mesi fa·discuss
I sort of wish there was a google "ad" search, where its like google search but only for ads, for the rare cases you want to buy something, and are looking through for a compatible product. Make advertisers differentiate by providing more information about their product to help me make a choice rather than shoving the product everywhere else hoping that I'll buy the thing out of fatigue
yarekt
·2 mesi fa·discuss
very cool, i’m in
yarekt
·2 mesi fa·discuss
Honestly I struggled to get further than 2nd paragraph. Other commenters here also have similar spectrum of opinions, far down the "Enjoy later" mindset
yarekt
·2 mesi fa·discuss
Jesus no, you could be dead or decrepit by 60. What’s wrong with finding happiness within your limits all the time?
yarekt
·3 mesi fa·discuss
except Juniors with Claude are now “seniors”.

In all seriousness even though deep down I know there’s no replacement for experience, I’ve seen a bunch of new people get impressively far with just hacking stuff together. And in the end if you’re making money with that for a long time, isn’t that all that matters to the companies?
yarekt
·3 mesi fa·discuss
I would say it’s more short term profit. They are happy to cut off the nose and put the rest in the backlog
yarekt
·3 mesi fa·discuss
You’ve missed the point of original article about the proxy for quality disappearing. LLMs are trained adversarially, if that’s a word. They are trained to not have any “tells”.

Working in a team isn’t adversarial, if i’m reviewing my colleague’s PR they are not trying to skirt around a feature, or cheat on tests.

I can tell when a human PR needs more in depth reviewing because small things may be out of place, a mutex that may not be needed, etc. I can ask them about it and their response will tell me whether they know what they are on about, or whether they need help in this area.

I’ve had LLM PRs be defended by their creator until proven to be a pile of bullshit, unfortunately only deep analysis gets you there
yarekt
·3 mesi fa·discuss
goddamnit, read for a solid 5 minutes until I realised the rest is paywalled, sigh. No pasty facts isn’t worth another subscription. I have many other interesting things to do
yarekt
·4 mesi fa·discuss
That’s a shame, though I do see that it is difficult to make any money from what it is. I’m glad they didn’t sell it to someone big for all the user’s data, though it is still early
yarekt
·6 mesi fa·discuss
In the UK I've not been able to find high wattage (10-20W) LED lightbulbs with high CRI, some don't even mention it in listings, let alone SSI, which I have never seen.

Where are you seeing these? Is this industrial/commercial suppliers?
yarekt
·6 mesi fa·discuss
except chess is a solved problem given enough compute power. This caused people to split into two camps, those that knew it was inevitable, and those that were shocked
yarekt
·6 mesi fa·discuss
Since no-one mentioned the Cynefin framework yet: Notice that there's a little fold at the bottom centre of the diagram: Chaotic (unknowable unknowns, etc) things will always resolve themselves to simple states.

Fires will eventually burn out, the result will be simple to understand. Simply your business won't exist anymore.

There are more nuanced examples but I believe the above explains the principle.

The Key is to handle things early, before the most probable/default resolution, if its one you're not happy with.
yarekt
·6 mesi fa·discuss
Nice, download a random binary off the internet and give it your AWS credentials.

Please people, inspect the source to your tools, or don't use them on production accounts.
yarekt
·8 mesi fa·discuss
No-one uses that original OOP at all, no-one sane anyway. The way its used now is for dependency injection. All your logic is in services that are injectable and unit tested. All your data is in simple immutable DTOs.

All the OOP tricks, classes, instances, interfaces, polymorphism, its all good for wiring up your logic, replacing bits at runtime. No-one actually models their domain with pure OOP. Urgh, that would be awful.

But also to echo other commenters, this isn't interesting insight...
yarekt
·8 mesi fa·discuss
not original commenter, but I have. Either through their manipulation, or just being in the same place, doing the same things. Didn't like them as a person, but they were decent to me, so some sort of reciprocation happened, didn't last though.
yarekt
·8 mesi fa·discuss
It's almost the same argument, but backwards: You think they are a good person, so you want them to do well. Because they are good, they also want you to do well. Same result, but intentions are backwards.