throwaway277432

They might be contraband in the future. In which case you could not rely on the archive to preserve them.

So the answer is both yes and no.

It doesn't matter, it's not a bug it's a feature. You don't have any power in this. Accept the thinking they give you.

That alone should disqualify it by itself.

I've run into both security and bio-refusals, it doesn't take much.

I recently got into an argument on HN when I said that: https://news.ycombinator.com/item?id=49002491

They already believe.

>the idea that an algorithm using matrix operations as a primitive cannot emulate "true" intelligence

I did not say that. Not even close.

In the words of Claude: I’ll end the conversation here. <end_conversation>

I'm not. But I'll concede you are possibly more literate - I won't dox myself on this account.

Congratulations.

If you honestly argue that LLM computation and human cognition are equivalent there is no further argument to be had - it's immediately a philosophical or worse a religious argument that cannot be won.

However, that we're even arguing on that level baffles me. How did this happen! They're glorified calculators.

They really marketed the hell (sorry) out of LLMs.

Yes, in the same way I like to kill the enemies in DOOM.

It's matrix multiplication. Absurd.

Claude (Opus 4.8) recently told me:

>I’d ask you to drop the abuse; (...) if it continues I’ll end the conversation.

After I'd used a couple of expletives. And yes it will emit a <end_conversation> token.

This is truly dystopian. It is NOT a person. What a response. I still cant believe it.

In my experience, these exec types are just completely oblivious.

They think the AI prose is great, they wouldn't think of proof-reading in detail and they absolutely think their time is more importantly spent elsewhere. Some of the stuff I see is just outright wrong - they don't care to get it right, they want to ship it.

Their peers on the other side don't care or notice either.

To be fair this is not a new phenomenon, it has always been like this. But the scope has changed.

They now think they don't need the engineers anymore and create "technical" documents themselves which are complete garbage and they also don't take any input on them. You can't argue with them because they don't understand what they wrote.

Pointing out errors is an exercise in futility as they actually believe the AI knows everything and is more correct than the engineers.

This assertiveness on their part is new and wasn't there before, and it's driving me mad.

We'll be out of work whilst bailing out these companies with printed money and another round of inflation. The middle class will cease to exist. But hey the stock market will show higher numbers so its all good.

>serious professional in this industry

As a serious professional in the industry - we're dinosaurs. Nobody cares anymore.

The kids are running the show and are making billions with stuff that doesn't work. But it makes money so nobody cares.

This is not a new phenomenon, it started years ago and really took off when JS became the new hotness. You could see it happening live, right here on HN. But the blast radius is massively increased now with AI and people are getting hurt. It's not funny.

The ship has sailed on rigor.

The sad thing is that this is not going to get better. The best we can hope for is slight improvements to agentic "engineering" practice with lots and lots of blog posts on HN written about how they are rediscovering basic engineering practices.

We (the dinosaurs) will roll our eyes while making a fraction of the money the kids are making.

And even if the whole AI ecosystem implodes (it won't) that would be a massive recession and certainly wouldn't make the remaining software engineering work more rigorous either.

As the Simpsons put it: "An I out of touch? No, it's the children who are wrong."

>Better pay.

>Once approved, our new bonus program will give every team member who isn’t on an incentive compensation plan or bonus plan today, the opportunity to earn a cash bonus based on their individual performance, targeting 10% of salary, awarded at their manager’s discretion.

LOL. So basically buckle up and do what you're told and grind. And hope your manager likes you or you'll get nothing.

What happens if you build a bridge and it breaks?

These people want to play god with our lives but at the same time move fast and break things. Look at software quality anywhere, it's a mess and only about to get much worse.

We should not let them. Jail time for anyone involved in any of the decision making process, applied at scale with the number of vehicles and deaths.

Why should the standards be any different? They want to change the status quo with tech only so they can get paid and extort us with yet more subscriptions.

AVs will never substantially reduce road deaths. They will optimize to just being slightly better than human, but fail in new and more unexpected ways. There is not enough incentive for them to make it safer.

No personal attacks please. They're just cogs in the machine.

The organization and process that enables it to get to this point is the problem. And that is MS, always has been.

To be fair, that makes it worse for MS, not better.

This should not be vibe-coded by someone who has absolutely no idea about any of these things.

>a project manager vibe-coded the change without thinking it through at all

The PMs vibe-coding and having no idea what they're doing isn't even the main issue (although it is pretty bad).

The main issue is: how are the actual engineers supposed to "review" the slop? They probably report to the same PM or are at below in the org chart and might be evaluated by them. Not just at MS, but any company.

Such a conflict of interest would be detrimental to quality anywhere. You wouldn't build a bridge like this, nor should you software.

Unironically yes.

I predict that costs will grow to 80% of what it would cost a human, across the board for everything AI can do.

"It's still cheaper than a human" they'll say. Loudly here on HN too.

Of course this will happen slowly, very slowly. Lets meet again in 10-20 years.

So? Everyone is saying to just look at the LLM outputs for PRs etc. and just ignore how it was created. We should apply that standard right here too.

This is Anthropics initial response, which they walked back ONLY because of the HN outrage. Without HN, that would've been tge official answer.

I'll judge them on that, thank you.

Having just worked my behind off for the last months to deliver on an impossible deadline, successfully: more bodies definitely would have helped.

Even just to keep the fluff off my back and to allow me to fully concentrate on what's important.

The situation will repeat itself in 6 months and I'm not going to do that again. Hiring now would fix that.

I'm not going to review it in full, sorry. Reviewing is so much more effort compared to producing something with AI. But don't let me deter you, keep on learning and keep on building.

I wish I had the possibilities to learn and build on such a large scale when I started out. AI is a blessing and a curse I guess.

My own early projects were most definitely crap, and I made the exact same mistakes in the past. Honestly my first attempts were surely worse. But my projects were also tiny and incomplete, so I never published them.

However: What little parts I did publish as open-source or PRs were meticulously reviewed before ever hitting send, and I knew these inside and out and they were as good as I could make it.

Vibe-coded software is complete but never as good as you could make it, so the effort in reviewing it is mostly wasted.

I guess what I'm trying to say is I'm a bit tired of seeing student-level projects on HN / Github cosplaying as production ready software built by an experienced engineer. It used to be possible to distinguish these from the README or other cues, but nowadays they all look professional and are unintentionally polluting the software space when I'm actually looking for something.

Please understand that this is not specifically directed at you, it's pent up frustration from reading HN projects over the last months. Old guy yelling at clouds.

>tell me if I earned your star

Since you asked: Not in a million years, no.

A bug of this type is either an honest typo or a sign that the author(s) don't take security seriously. Even if it were a typo, any serious author would've put a large FIXME right there when adding that line disabling verification. I know I would. In any case a huge red flag for a mitm tool.

Seeing that it's vibe coded leads me believe it's due to AI slop, not a simple typo from debugging.

I got a major reprimand because I answered too many questions posted in the public channel. All in my area of expertise, mostly after hours.

At first they said it was "great". But it soon turned sour and resulted in "it seems like you spend too much time answering questions", and I should "focus" and "free up" that time to work on my assigned tasks.

Well, I don't answer anything anymore. In fact nobody does. It used to be that you got precise technical answers from someone directly working on the tool or problem you asked about. The previous CEO would sometime even answer themself. Not anymore.

Now people ask, but nobody answers. The rest has devolved into LinkedIn style self-promotions and announcements.

Lots of people haven't had to actually restore their data. Somehow it has good marketing. I used it for a while and was not impressed. Random Python errors, requires too much scripting, and at least on my data terrible restore speed.

I followed development on Github and what I saw in terms of fixes and commits gave me pause. Not how I like my critical backup software written.

I now use restic and sleep much better.

Don't ever travel, never change anything related to billing except to update your cards before they expire. Don't change your name, email adresses or lose access to your phone number, and as we know now also don't ask support.

Then don't use any uncommon tools, e.g. ones associated with 'hacking', or store any copyrighted files in their cloud.

If there's any issue or error with logins etc., don't retry too quickly or too often or that in itself will be suspicious. Wait a day between requests, and double-check everything before retrying. Do not retry from a different IP or worse a VPN, or that will also be suspicious.

That should just about cover the bases for most providers.

Yes, it's insane and obviously you still need a backup of all your stuff just in case.

> "What do you think about the comments of user XYZ"

Wow that is really scary. Never did I ever think someone would actually go through all my old comments, analyze them in detail and then judge me based on them (my real account, not this throwaway).

Yes I knew it would be theoretically possible, but you'd have to be a total stalker and real creep to actually do it. Now anyone with an LLM can just do it without a second thought.

And it'll only get worse from here on. I'm sure there is at least 1 comment somewhere on the internet by me where I wasn't too nice, or a like / upvote on a questionable opinion or something.

If it's in any way connectable to me future AI tech is going to find it. Probably even across accounts, matching writing styles and whatnot.

I seriously think I'm going to stop posting on the internet for good.

>chore: change to MIT license

What does "chore" mean in this context? Is the license just leftover from some MS open source template? If so there is perhaps some leeway, and the author maybe just didn't realize he needed to use the original MIT license file including the notices and not just a template one grabbed from the internet.

Any other explanation for such a "relicensing" would be extremely worrisome.

I've seen plenty of both. I've added one good example in my other comment. But it certainly depends on the community and programming language as to how serious licensing is treated.

But yes, many people are not complying with the license literally, and it's frustrating to see. I know it basically doesn't matter unless you go to court over it, but still it irks me and screams a sort of carelessness about the rules and social contract.

Sorry for criticising your reading comprehension, I did not mean it as a personal insult.

It's just that I see these types of responses so often, basically every time any licensing question comes up. Twice in this thread. And all that's required is to just read the very short and basic MIT license text itself, no lawyering required.

I can understand the native speaker part, but just know that I myself am not a native speaker either. But I understand that's a huge barrier.

But even native speakers on HN with serious software engineering jobs and skill don't understand it, or don't want to understand. I think it's a bit like when people see math proofs, they mentally just skip over it.

That's the part that continues to amaze me.

The easiest way to do it is to add your own copyright line above the original LICENSE copyright line.

That way anyone touching the project can just add their own line on top.

Done.

EDIT: Example: https://github.com/gcnlab/gitea-fork/blob/main/LICENSE

A more complicated way to do it is to add a folder that contains the original LICENSE file or files. Sometimes there is more than one license, or the license texts differ. In that case, you must preserve all the different variants, even if they all call themselves MIT.

Then, you can optionally add your additional own LICENSE file * only iff* it is compatible with all existing LICENSES. In the case of the MIT license, you may relicense, sublicense, or use a different license in addition, provided it is MIT-compatible. With e.g. GPL you can't. Note that you still have to preserve all the original LICENSE files in the repo.

No!

Once you change the copyright line, you no longer include "the above copyright notice". At that point you're violating the license.

You are also not allowed to change the copyright notice or license text in any way (you may however add to the license, which is a loophole other licenses such as GPL fix.)

Substantial is subject to (legal) debate as the Oracle vs. MS case has shown. Whole functions or large parts of files however should always be considered substantial, as the software would otherwise not work.

I'm seriously flabbergasted at how bad reading comprehension seems to be among coders.

It's been happening for years. They don't care.

See e.g. Microsoft forked MIT licensed repo and changed the copyright (1326 points, 538 comments, Dec 25, 2021) https://news.ycombinator.com/item?id=29683471

That time their excuse was "our automation was wrong". I wonder what it'll be this time.

It will probably be fixed and they'll say sorry and so on but MS doesn't really care about this at their core.

Embrace, extend, extinguish is what works, and they know it.

I was briefly diagnosed with "99.9% sure it's cancer" before it turned out to be benign. Say about 2-3 weeks.

In those few weeks my main regrets were a) not having done many of the things on my bucket list, and b) not having children or not going to be live long enough to see them grow up.

I'm someone with recurring nightmare about career goals and such. However at that time, work only crossed my mind briefly and was easily dismissed.