veggieroll

Yes! Can confirm: I went to the doctor when I was very overweight (on the verge of obese BMI-wise) and he pulled this line. He was a nice guy. But, I couldn't get out of there fast enough. Why go to a doctor that won't even acknowledge one of the most important things you can do for your health?

Ok, do rent next.

Paying rent doesn't count towards your credit score (at least it didn't a decade ago when I was getting my mortgage). And it's a glaring signal of how the system isn't built for what it claims: evaluating your ability to pay obligations on time. No, the system is built to trap people in poverty and enrich the rentier class.

Yo, this installs npm packages at runtime. Very not cool IMO. You should disclose this prominently in the README.

This is a nice project idea. But, you should use a Python WASM interpreter to run the PostgreSQL WASM.

I have yet to see a proposal that retains what I love about the status quo: conscientious error handling.

The language’s status quo forces everyone to think about errors more deeply than in other languages and acknowledges that the error case is as critical and worthy of the programmer’s attention.

Error handling is one of my favorite parts of Go. The haters can rip `if err != nil { return fmt.Errorf("error doing thing: %w", err) }` from my cold dead hands.

Ah! That makes more sense. Thank you for clarifying. Great to see you got so much activity on your HN thread here.

I think there's a really interesting opportunity for a synthesis of the classic NPC dialog menu and a fully freeform LLM character.

Namely, the dialog would still be fixed, where there's a pre-defined flow of the conversation and fixed set of facts to deliver to the player. But the LLM could generate variations on it each time so it's never exactly the same twice. And it could add more character so the NPC gets frustrated if you keep asking it over and over. Or, it tries to dumb it down for you. Or, it gets sick of you and just tells you point blank: Look, you need to know XYZ. That's all I have for you.

Namely, the dialog would still be fixed, where there's a pre-defined flow of the conversation and fixed set of facts to deliver to the player. But the LLM could generate variations on it each time so it's never exactly the same twice. And it could add more character so the NPC gets frustrated if you keep asking it over and over. Or, it tries to dumb it down for you. Or, it gets sick of you and just tells you point blank: Look, you need to know XYZ. That's all I have for you.

Thanks for the explanation! I'm learning and and think this would be a good next project for me to try, especially since I have a real world use case in mind with a similar amount of data available.

In particular, I'm not very familiar with reinforcement learning and am not sure how you use the embeddings from MiniLM2 as a reward function. (Edit: maybe this is the jaccard similarity?)

I'd really appreciate it if you were open to posting scripts! I see a few snippets around and could probably cobble something together after a while. But, it's cool to see something already working to make sure I'm not getting too far off into left field.

Have you written anywhere in detail on how you gathered your dataset and trained the finetune? I have a few use cases that are like this, but I'm not sure where to start.

You're right that the power dynamic here is really bleak. But there are still some things an employee can do individually or in a group:

- Coffee Badging: where you only come in for coffee in the morning and then leave

- Buddy Badging: team up with coworkers badge to scan each other in on alternating days, weeks, whatever

- Quiet Quitting / Lying Flat: aka. chilling on performance while you wait to get fired or look for another job

- Malicious compliance: make a lot of noise, microwave fish in the break room, or anything else you can do to terrorize decision-makers

- And finally the going for the jugular....try to organize your coworkers into collective action, ultimately working towards a union

GIMP is a load bearing name as well.

Cheers on sticking to your guns. Prominent displays of deviance as a poison pill is truly one of the few ways subcultures have to resist mainstream exploitation.

How do you handle temp file space, timezone data, and other things that a minimal image provide?

Yes!! This is my preferred way to enable network printing and scanning. I've done this for about 5 years.

I've always hated having printers on my network, because vendor software is absolutely terrible. Completely insecure and open to who-knows-what through automatic updates.

I use a raspberry Pi and it works great even for non-technical family members. Printing from your phone is also really nice.

The functionality is about the same as the vendor printers. But the peace of mind is amazing. 100% assurance that there's nothing weird going on, it's secure, and there's no account and stupid cloud.

So much fun! Great work.

Chrząszczyżewoszyce powiat Łękołody

> Brzęczyszczykiewicz

For anyone not familiar, this is a Polish joke. https://www.youtube.com/watch?v=AfKZclMWS1U

Not only that, but he actively made it worse by protecting abusers. McCarrick, Rupnik, Zanchetta, and many more. It really makes me sick to see so many people speaking well of Pope Francis. He is an enabler and a vindictive hypocrite.

I've been using my passport to fly ever since 9/11, because I saw where the wind was blowing. We've been falling into a "papers please" state for a long time.

My dream is that the government abolishes the TSA and we return to pre-9/11 security. Recently, this was proposed in Congress. [1]

[1]: https://www.congress.gov/bill/119th-congress/senate-bill/118...

No, indeed I found a way to skip physics in high school (though this wasn't really why). But, I was interested in Biology, taking almost enough for a minor in it in college.

I'm a self-taught dev now. And, that fits really well for me, despite being completely unrelated to my college degrees. I work mostly with other self-taught, passionate about software people. And I'm loving that.

But, I do have very strong opinions on institutions and pedagogy. I've gotten into some pretty epic arguments about it with my wife, who is a music teacher. And, her experience has been so completely opposite of mine.

From the way she tells it, classical music seems to be the ultimate discipline where structured education is paramount. And, I have such a negative opinion of traditional methods that it's caused some friction.

> scepticism of the validity of learning institutions

Can confirm, this is solidly wedged into my opinions now. There were a lot of other experiences after this to compound that feeling.

In High School, I started looking aggressively for a less traditional path and fortunately found one. It really saved me, because I was forced back into the traditional environment in Senior year of high school, and my grades tanked from top-of-class to "you might need summer school to graduate" level

Things got a lot better in college, because that experience (among others) helped me effectively navigate the institution, jump directly into more advanced coursework, and earn more freedom to study things that were interesting to me.

I did get a job in my field out of college. So, my college pedigree was useful practically (though not really any knowledge I got there). But, I'm self-taught dev now, which is an amazing fit for my experience and attitude.

I can totally relate. I had the same experience in grade school science class, where the teacher assigned an experiment with a suggested solution and an invitation to come up with your own method.

I was the only person in class that chose to do my own method. And, it didn't work because I didn't account for an environmental difference between my house and the school classroom. And, he gave me a failing grade.

It really killed my interest in physics for a long time. I focused on biology from then through college.

Ultimately, the problem was that he didn't make clear that the only thing that we were being graded on was accuracy, not experimental methods or precision. (My solution was precise, but inaccurate; whereas the standard solution was accurate but imprecise) Also, it's possible everyone else in class knew the culture of the school, and I didn't because it was my first year there. So, I didn't realize that they didn't value creativity in the way I was used to.

I didn't say no-one gets upset by the name. I said that they're being silly gooses.

If your department wouldn't use free software because of the name, that's very much a them problem. And it's sad that they're so sensitive.

If you don't like the name, you don't have to use the software. Or fork it can call it something more anodyne to your sensibilities.

But no. You want everyone to cater to sensitive people. And thus, myself and others are here to say your comment is annoying when this comes up every thread that GIMP is mentioned.

Has anyone actually tried to use AI in a live environment to order takeout or make a reservation by phone? I am really tempted to try, but feel like it's deeply violating an unwritten rule of human decency to do this on an unsuspecting local restaurant.

If I was a restaurant worker, there's a pretty high likelihood I'd just hang up on a call like this or try to call back and reach a human.

p̶i̶c̶s̶ model weights or it didn't happen.

Fork it then.

Literally, get over it. Even my (extremely conservative) church group doesn't have an issue with it. 1) because most people don't know BDSM terminology and 2) gimp has other connotations so it's not even obvious that was intended and 3) most people aren't terminally online and 100% don't care as long as the program works.

I've never had a problem sharing it. And no one outside tech has ever brought up BDSM when I tell them about the program or help them install it.

That's probably true. But offending people with prudish attitudes or without a sense of humor is a pretty common old-school hacker goal.

It serves a valuable community function by protecting the community from corporate interests, non-technical bike-shedding, and overload by "bug" reports that are actually just the user not being technical enough to use the tool properly.

Edit: Also FWIW, I've had no problem introducing GIMP to church groups and in various professional settings.

Double edit: since you mention schools, I should also add that I taught my wife GIMP and she had no problem sharing it in her school (US public high school).

Absolutely! But the question is whether the next step-change in intelligence is just around the corner (in which case, this legal speedbump might spur innovation). Or, will the next revolution take a while.

There's enough money in the market to fund a lot of research into totally novel underlying methods. But if it takes too long, investors and lawmakers will just move to make what already works legal, because it is useful.

I mean, you're right in the abstract. If you train an LLM in a void and never do anything with the model, sure.

But that's not what anyone is doing. People train models so that someone can actually use them. So I'm not sure how your comment is helpful other than to point out that distinction (which doesn't make much difference in this case specifically or how copyright applies for LLM's in general)

> Thomson Reuters prevailed on two of the four factors, but Bibas described the fourth as the most important, and ruled that Ross “meant to compete with Westlaw by developing a market substitute.”

Yep. That's what people have been saying all along. If the intent is to substitute the original, then copying is not fair use.

But the problem is that the current method for training requires this volume of data. So the models are legitimately not viable without massive copyright infringement.

It'll be interesting to see how a defendant with a larger wallet will fare. But this doesn't look good.

Though big-picture, it seems to me that the money-ed interests will ensure that even if the current legal landscape doesn't allow LLM's to exist, then they will lobby HARD until it is allowed. This is inevitable now that it's at least partially framed in national security terms.

But I'd hope that this means there is a chance that if models have to train on all of human content, the weights will be available for free to all humans. If it requires massive copyright infringement on our content, we should all have an ownership stake in the resulting models.