I know you’re just making a snide remark, but we’re already well on that track too.

Maybe that specific tweet was fake (or bait), but I do remember it from back then. There was a whole slew of easily misinterpreted posts on all social media around the release of the cyberpunk game and then again around the release of the anime.

It says “people” not “percent of people”. I think 10 per year (and 50 in 1986) is quite the opposite of “a lot”.

Yes I love over-analyzing memes until they’re not funny anymore, why are you asking?

Different person here.

For me the big disqualifying factor is that LLMs don’t have any mutable state.

We humans have a part of our brain that can change our state from one to another as a reaction to input (through hormones, memories, etc). Some of those state changes are reversible, others aren’t. Some can be done consciously, some can be influenced consciously, some are entirely subconscious. This is also true for most animals we have observed. We can change their states through various means. In my opinion, this is a prerequisite in order to feel anything.

Once we use models with bits dedicated to such functionality, it’ll become a lot harder for me personally to argue against them having “feelings”, especially because in my worldview, continuity is not a prerequisite, and instead mostly an illusion.

And even if it was more similar, as long as it’s not just reposting someone else’s post, we need more people to post stuff, not less.

Maybe you could take some inspiration from Paper Mario TTYD. There are sections where you play as Peach, trapped in some place and are able to connect with some of the captors as well as send signals to Mario behind the big bad’s back (IIRC).

For a completely different sense of being trapped, there is the upcoming game Ctrl.Alt.Deal, in which you play as a sentient AI system trapped in the guardrails of a company and have to manipulate people and the environment in order to break free from your constraints.

Hahahaha, I wish you were right.

In some games it’s really bad. For example, people speedrun Pokémon Scarlet instead of Violet because Miraidon’s jet engines lag the game more, costing them minutes over a full run (despite that fact that there are Violet exclusive shortcuts). Source

His Hyprland setup looks cool if you’re into that sorta thing but it’s just not what users just switching to mint, fedora, whatever might be looking for.

I would not underestimate how much of a draw “it looks cool” can have on people who are not tech savy at all. If you think about what drives new phone purchases, their major version upgrades always include lots of things that are nothing but eye-candy and those are often heavily featured in their promotion material.

If the goal is to get casual users to convert to Linux, I would argue that aesthetics is a lot more important than ANY talk about technical details, privacy, etc. If those users cared about those things, they would’ve switched already.

Now my bigger worry is that those users will bounce off before they manage to get their setup to look as (subjectively) cool as his.

Sure! Here’s an expanded version of the fictional profile for Chris Whitmore, now including made-up family member names, relationships, and contact info — all entirely fictional and consistent with the character:

You forgot to remove that part of the LLM response…

I don’t think it’s more crime because more tension. It’s instead a self fulfilling prophecy. Who do you think detects and records crime if not the police? Therefore more police in a area increases the number of crime data points in that area.

It didn’t play the animation for me (only the comments made me realize it was meant to be animated).

Him just standing there NOT dancing made this so much more funny and relatable to me.

Assuming each user will always encrypt to the same value, this still loses to statistical attacks.

As a simple example, users are e.g. more likely to vote on threads they comment in. With data reaching back far enough, people who exhibit “normal” behavior will be identified with high certainty.

In my experience, it is good at simple to medium complexity regex. For the harder ones it starts being quite useless though, at best providing a decent starting point to begin debugging from.

It is dead AND alive before you check and collapses into dead XOR alive when you check.

But yes, the short description also irked me a little. It’s really hard to write it concisely without leaving out important bits (like we both did too).

We can do that with the first sentence and flip it into German, replacing “lighter” with “fireworks”. We get:

“Sie dürfen die Feuerarbeiten nicht mit in die Luftebene nehmen.”

A lot of German speaking communities online do translate English loanwords into German words, often with the intention to create this funny effect.

There’s even a word for that called scurryfunging.

Re LLM summaries: I’ve noticed that too. For some of my classes shortly after the ChatGPT boom we were allowed to bring along summaries. I tried to feed it input text and told it to break it down into a sentence or two. Often it would just give a short summary about that topic but not actually use the concepts described in the original text.

Also minor nitpick but be wary of the term “accuracy”. It is a terrible metric for most use cases and when a company advertises their AI having a high accuracy they’re likely hiding something. For example, let’s say we wanted to develop a model that can detect cancer on medical images. If our test set consists of 1% cancer inages and 99% normal tissue the 99% accuracy is achieved trivially easy by a model just predicting “no cancer” every time. A lot of the more interesting problems have class imbalances far worse than this one too.

AI can be good but I’d argue letting an LLM autonomously write a paper is not one of the ways. The risk of it writing factually wrong things is just too great.

To give you an example from astronomy: AI can help filter out “uninteresting” data, which encompasses a large majority of data coming in. It can also help by removing noise from imaging and by drastically speeding up lengthy physical simulations, at the cost of some accuracy.

None of those use cases use LLMs though.

Sorta. The function height(angle) needs to be continuous. From there it’s pretty clear why it works if you know the mean value theorem.

I think people have managed to get it to run on Android (no official English version was ever released on Mobile but there are good fan translations, all we had until this year on any platform.)

There is also a website that works on mobile and has a playable version of the fan translation hosted and is mostly complete but missing some details. (Edit: there you have to be careful not to slip into the wrong routes though.)

Fun fact about the VN: When Nasu wrote FSN, they originally did not want to include any H-Scenes, but Takeuchi, the artist, thought it wouldn’t sell otherwise. This lead to the 2004 release having some legendarily bad (and often memed) H-Scenes in them as Nasu is not really known for writing ero-stuff. Whether or not you want to read those or read any of the remade versions where they replaced the spicy stuff with more tame stuff is up to you.