People developing local models generally have to know what they’re doing on some level, and I’d hope they understand what their model is and isn’t appropriate for by the time they have it up and running.

Don’t get me wrong, I think LLMs can be useful in some scenarios, and can be a worthwhile jumping off point for someone who doesn’t know where to start. My concern is with the cultural issues and expectations/hype surrounding “AI”. With how the tech is marketed, it’s pretty clear that the end goal is for someone to use the product as a virtual assistant endpoint for as much information (and interaction) as it’s possible to shoehorn through.

Addendum: local models can help with this issue, as they’re on one’s own hardware, but still need to be deployed and used with reasonable expectations: that it is a fallible aggregation tool, not to be taken as an authority in any way, shape, or form.

On the whole, maybe LLMs do make these subjects more accessible in a way that’s a net-positive, but there are a lot of monied interests that make positive, transparent design choices unlikely. The companies that create and tweak these generalized models want to make a return in the long run. Consequently, they have deliberately made their products speak in authoritative, neutral tones to make them seem more correct, unbiased and trustworthy to people.

The problem is that LLMs ‘hallucinate’ details as an unavoidable consequence of their design. People can tell untruths as well, but if a person lies or misspeaks about a scientific study, they can be called out on it. An LLM cannot be held accountable in the same way, as it’s essentially a complex statistical prediction algorithm. Non-savvy users can easily be fed misinfo straight from the tap, and bad actors can easily generate correct-sounding misinformation to deliberately try and sway others.

ChatGPT completely fabricating authors, titles, and even (fake) links to studies is a known problem. Far too often, unsuspecting users take its output at face value and believe it to be correct because it sounds correct. This is bad, and part of the issue is marketing these models as though they’re intelligent. They’re very good at generating plausible responses, but this should never be construed as them being good at generating correct ones.

Ideally, I agree wholeheartedly. American gun culture multiplies the damage of every other issue we have by a lot

One or more parents in denial that there’s anything wrong with their kids and/or the idea they need to take gun storage seriously? That’s the first thing that comes to mind, and it’s not uncommon in the US. Especially when you consider that a lot of gun rhetoric revolves around self defense in an emergency/home invasion, not having at least one gun readily available defeats the main purpose in their minds.

edit: meant to respond to django@discuss.tchncs.de

90 days to cycle private tokens/keys?

Days of the week in Spanish follow almost the exact same pattern as they do in English, with the exception of the weekend.

Specifically, Friday in Japanese translates to Gold Day, which also has a lot of cultural connotations with it being the lucky day of the week

I believe the wording is “drinks which burn the throat” which naturally means:

• ❌ Coffee
• ❌ Alcoholic drinks
• ❌ Coca Cola
• ❌ Hot tea
• ❌ Chai lattes
• ✅ Sprite, other non-caffeinated soft drinks
• ✅ Hot Chocolate
• ✅ Kombucha
• ✅ Energy drinks???
• ✅ Herbal Tea (even while hot, but mostly if you’re sick as a home remedy)

Most of the focus is interpreted as “contains caffeine and/or alcohol” but the wording is vague enough that it leaves for a lot of weird wiggle room people try to argue (based on convenience usually). It’s quite silly

~Babe, wake up! New feminization technique just dropped~

Life is suffering. Once you accept that fact wholly, you may ascend

And road design factors into the average driver’s speed far more than a sign with with a number on it

Driving is more fun when there are more viable alternatives. I don’t like driving, but it’s my only real choice where I live so I do it begrudgingly, and you have to share the road with me. Think of all the people who don’t want to drive (on account of it being dangerous, costly and/or mentally taxing) suddenly not being in cars, and how much traffic that would free up for you to zip around instead!

Also, calling a public service “bankrupt” is really weird to me. How many tax dollars are we spending on public highways and freeways again? Do suburbs, which are designed to be car-dependent, provide a net gain or net cost in tax revenue to cities?

From my experience, the pineapple pizza haters are way more toxic than the pineapple pizza enjoyers.

Women are so good and you can also be the best 😊

Women are in my house 🏠

Women are often a lot better for me 😜

(It suggested the emoji as well)

Cool, now explain that to a class of 7th graders for me

edit: raised grade, it may been set a low but it varies. I think most kids start to learn this stuff in/around middle school

edit 2: also mirrored objects are generally considered similar so that’s fine

Yes, and introductory geometry courses teach students how to do uniform scaling far before they teach them axis-based because it’s better illustrative of the concepts of similarity and congruence

The scaled down rectangle should be narrower; it’s not scaled in this diagram, it’s squished.

(Yes I know you can ‘scale’ objects on one axis but that’s usually not how it’s taught on an introductory level. Standard scaling assumes object similarity, which is not present in the diagram’s ‘scaled’ rectangle.)

That’s really good to know, and not how I thought the system worked previously. I thought instances were responsible for all vote aggregation and simply reported totals to each other at regular intervals, plus submitting comments/edits from users which are more obviously public

Y’know, that’s fair. I think I misspoke, and meant to say that the admins of your instance can see your IP but not the admins of another (assuming you’re not self hosting on your home PC without a VPN), but I’m not 100% sure that’s true because I’ve never looked at the protocol.

If every interaction is already public on the backend/API level, then simply not showing the info to users is just a transparency issue.

The more I’m thinking about this, the more I believe it’s a cultural/expectations thing. On websites like Tumblr, all of your reblogs and likes are public info, but it’s very up front about that. Social media like Facebook, IG, and sites like Discord, it’s the same; you can look through the list of everyone who reacted.

Data is not suddenly public just because some people have access to it. Data is public when it’s available for anyone to look at. Privacy is almost always going to be a trust issue on some level, and very few things are possible to do truly anonymously. Some data will always be available to someone in a position where it’s possible to abuse. Instance admins can see your IP address. Should that be available for everyone to see?