TehPers

I bought a license many, many years ago and loved SmartGit. I just use the cli now, but if you’re looking for a GUI, it’s a great choice.

It’s not really (which says more about corporations than anything).

Yep! You would need not only an AI superintelligence capable of reflecting and adapting, but legislation which holds liable those superintelligences and grants them the rights and obligations of a human. Because there is no concept of reward of punishment to a LLM, they can never be replacements for people.

There is a fundamental limitation of all LLMs that prevents it from doing as much as you might think, regardless of how accurate they are (and they are not):

LLMs cannot take liability. When they make mistakes, they cannot take responsibility for those mistakes. The person who used the LLM will always be liable instead.

So any automation as a result of LLMs removing jobs will end up punting that liability to the next person up the chain. Management will literally have nobody to blame but themselves, and that’s their worst nightmare.

Anyway, this is of course assuming capabilities that don’t exist.

The conclusion of this experiment is objectively wrong when generalized. At work, to my disappointment, we have been trying for years to make this work, and it has been failure after failure (and I wish we’d just stop, but eventually we moved to more useful stuff like building tools adjacent to the problem, which is honestly the only reason I stuck around).

There are a couple reasons why this problem cannot succeed:

1. The outputs of LLMs are nondeterministic. Most problems require determinism. For example, REST API standards require idempotency from some kinds of requests, and a LLM without a fixed seed and a temperature of 0 will return different responses at least some of the time.
2. Most real-world problems are not simple input-output machines. When calling, let’s say for example, an API to post a message to Lemmy, that endpoint does a lot of work. It needs to store the message in the darabase, federate the message, and verify that the message is safe. It also needs to validate the user’s credential before all of this, and it needs to record telemetry for observability purposes. LLMs are not able to do all this. They might, if you’re really lucky, be able to generate code that does this, but a single LLM call can’t do it by itself.
3. Some real world problems operate on unbounded input sizes. Context sizes are constrained and as currently designed cannot handle unbounded inputs. See signal processing for an example of this, and for an example of a problem a LLM cannot solve because it cannot receive the input.
4. LLM outputs cannot be deterministically improved. You can make changes to prompts and so on but the output will not monotonically improve when doing this. Improving one result often means sacrificing another result.
5. The kinds of models you want to run are not in your control. Using Claude? K Anthropic updated the model and now your outputs all changed and you need to update your prompts again. This fucked us over many times.

The list keeps going on. My suggestion? Just don’t. You’ll spend less time implementing the thing than trying to get an LLM to do it. You’ll save operating expenses. You’ll be less of an asshole.

Graham Platner isn’t an elected official, nor are any of the three legislators named Graham Platner, nor are any of Louis Rossman’s other names Graham Platner.

Please, if you’re going to troll, make a better attempt at it. Or even better, take it somewhere else.

He literally put out a video not that long ago praising three democratic legislators for proposing a right to repair bill. Weird impression.

By “crackers” I mean “black hat hackers”.

Ok I just misread it then, sorry!

What they’re doing should be outright illegal in most countries; it’s equivalent to changing a contract unilaterally after both parties signed it.

Update to [COMPANY NAME]'s Policies

Yes, this should be illegal, but it’s already common practice. I’m just hoping that enough of this will eventually get people to stop buying these products, and hopefully we can start seeing some real legislation against it in some countries.

Additionally, I’d strongly advise against buying any sort of “smart” device, unless you’re pretty sure the benefits of connecting your toaster to the internet outweighs all the risks.

This should be obvious at this point. “Smart” just means “internet-connected”, and we already know what happens to every device that connects to a remote server during regular operation: telemetry (and not the nice debugging kind but the “what do you use” kind), and advertisements.

Including corporations and crackers

The “crackers” part of this confuses me. Samsung is a Korean company. The chairman’s name is Lee Jae-yong (이재용). Samsung NA’s CEO is Yoonie Joung. Maybe I’m misreading this?

Lol we did something like this at work once and it was pretty fun.

Here’s an idea: it’s a royal pain in the ass to transfer Firefox data from one mobile device to another without using their cloud sync feature. I’d love a way to do that while keeping the data local only.

It will probably end up running better for them anyway.

Seriously, I use Office at work, and I wonder everyday why anyone subjects themselves to it. Then I subject myself to it because that’s what I’m supposed to use. But I sure as hell don’t at home.

It sounded from the PSF article that the biggest reason they rejected it, aside from the ethical concerns, was that they could be asked to give the money back if they are found to be violating that term anywhere and at any time. Even if they didn’t care about the term regarding DEI, the risk of being liable for $1.5m to the US government all because an orange blob didn’t like the logo is too great of a risk.

I mean, I find TypeScript fun to write. The only thing I really dislike about it is configuring the tools (tsc, eslint, etc). It’s a great language when everything’s setup and you disallow all the ugly JSisms with your linter and tsc.

Used Claude 4 for something at work (not much of a choice here and that team said they generate all their code). It’s sycophantic af. Between “you’re absolutely right” and it confidently making stuff up, I’ve wasted 20 minutes and an unknown number of tokens on it generating a non-functional unit test and then failing to solve the type errors and eslint errors.

There are some times it was faster to use, sure, but only because I don’t have the time to learn the APIs myself due to having to deliver an entire feature in a week by myself (rest of the team doesn’t know frontend) and other shitty high level management decisions.

At the end of the day, I learned nothing by using it, the tests pass but I have no clue if they test the right edge cases, and I guess I get to merge my code and never work on this project again.

I was being ridiculed in the past and called a slop-generator

I can only imagine why. Surely it’s unrelated to this?

I’ve completely moved to codex cli as daily driver. I run between 3-8 in parallel in a 3x3 terminal grid

Nah, couldn’t be.

Priority handles will be free and “often include full names, multi-word phrases, or alphanumeric combinations.” Rare handles, on the other hand, will be a paid option, and “may be priced anywhere from $2,500 to over seven figures, depending on demand and uniqueness.”

Ok so if it’s a “high demand” handle then I guess they bend you over for it. Domains are like that too, but it isn’t exactly something to strive for.

But more importantly, if you ever downgrade your X subscription, your account will revert back to your original username, and you’ll lose access to the one you snagged through the marketplace.

I really hope someone pays 7 figures for one only to get fucked by this. I want to see it so badly.

On some teams, the backlog is just a way to track tech debt. Backlogs have a habit of growing indefinitely making it so lower priority tasks never get done.

Eithet way, tech debt is generally a bad thing. You’re just weighing it against other tasks, which are almost always more important.

Nah, it’s futuristic. It’s the embodiment of the future where car manufacturers sell you rust buckets full of tracking and make you pay way too much for it.

I’ve seen some wrapped ones that look neat though, and anyone who owns a Cybertruck and isn’t wrapping it is just wasting more money than they already did buying it.

Windows 11 also comes with some new, great features that you’ll miss by not upgrading! Your windows will sometimes decide not to repaint their surfaces, especially under load or while screen sharing. You’ll have a OneDrive ad in the main settings page. Copilot and Teams will randomly install themselves when installing your forced updates. And we can’t forget about all the work being done to ensure you have no choice but to link your account with a MS account!

Look, it used to be tolerable. The bugs made me install Linux, and I haven’t missed it at all.

Right to repair does not match well with right wing politics (TL;DR AuthRight need control and in lib-right absolute capitalism having reparable stuff is a surefire to kill your company ) but it’s a murky and difficult subject so I understand why it’s not mentioned.

While I agree, especially around farming, right to repair is a massive topic and advocated for strongly. It’s weird that they’d then advocate against it with their other views, but logic hasn’t existed in politics for longer than I’ve been alive, so yeah.