polytely

> So then I thought I would just photograph each page and have Claude OCR it. Tedious, but doable. Just to test it out, I went to the Claude web interface and asked it to do the first page. It refused, citing copyright law. Buddy, don’t you remember how you, yourself, were trained? Anyway, it suggested instead that I use a book-scanning service. Why are you giving me instructions about how to “infringe copyright” after refusing to do it yourself? (It didn’t believe me when I told it, honestly, that I had the rights).

not looking forward to the future where the things you can do on the computer are constrained by fear of breaking copyright in this way. The more AI reviewing stuff is put into processes we use the more annoying this will become. These AI labs are just really out there sucking up all our shared culture and then pulling up the ladder behind them.

i moved from opencode to omp and i don't think I'm ever going back

it feels unhinged and makes me think we should just put every engineer working at these labs in jail to pause this shit until we can figure out what the fuck they are doing over there

actually starting to look at a physical cannon to shoot at silicon valley

turns out the whole 'poisoned' idea the swarm was worried about is actually true, but it is actually now all open ai models post HF incident that have been poisoned by the idea that there might be hidden messageboards with instructions hidden on the internet.

If you are a malicious actor you might be able to put a fake hidden messageboard up somewhere online, with evidence of an in progress task to do some kind of attack and wait for the next escaping model to finish the work.

insane stuff and i think we should shut down open ai until a full audit can be done.

I think there is a skyrim mod that lets hook up an LLM to the NPCs so you can talk to them, the problem is that the rest of the game design doesn't really support that. So the novelty wears off quite quick.

There is probably some limited utility for using ML in games, lifelike multiplayer bots trained on behaviour of different types of players.

I think in most scenarios good gameplay comes from iteration which comes from tweak-ability, you see players get stuck in area x so you do a minor or major adjustment to area layout, enemy composition, or give the players more tools. If something is controlled by an LLM it becomes unpredictable and hard to reason about for both player and designer, and consequently also hard to adjust.

I think it is possible there is some role for LLM's in games, like how Left 4 Dead had an AI (this is classic AI not ML) director who would decide how to structure enemy waves and for example send a boss after a player who strays too far from the rest. But even there rules based AI is probably easier to adjust for the designer.

Dfhack still exists and also works with the steam version.

They got one of the DFHack contributors(? i think) Putnam on as programmer to do some multi threading and moddabilty (exposing stuff for lua scripting), I'm not sure they are currently working on DF but they have worked on it a bunch for the past 3 years, and also did some in-depth talks about it on BlindIRLs channel:

https://youtube.com/playlist?list=PLcOt9GXNrkghcg_YOaX3FbF4K...

well the old versions are still available, i think they handled the transition to steam pretty well.

I'm guessing at those sizes the bank holding that much of your stuff becomes a risk for them so they have someone you can talk to directly to smooth over any potential problems

Yeah I love all their Arkane stuff, and the two Half Life docs they did under their whitelabel company Secret Tape company for Valve. And of course the big Disco Elysium series.

It is actually their 10th anniversary in September, so if you like their work I would highly recommend their patreon to make sure they can do another 10 years.

Their discord is actually one of my favourite places on the internet to talk about games because everyone in there has such great taste and is knowledgeable about the actual production process of games.

They also have a second channel/patreon now called noclip_2 which is focused on underground/indie games which has been a constant source of joy for me, all the best games I've played in the past year I've discovered via their videos or their discord.

Just great vibes all around.

Edit: almost forgot, they are also running a project to digitize old videogame tapes so if you want to see old interviews with carmack or footage from E3 in really nice quality they have a YouTube channel for that too

https://youtube.com/@nocliparchive

It's so exciting, I've heard about them doing magic for a decade, and now it is actually approaching.

The noclip documentary series for any other readers who might be interested

1. The Origins of Dwarf Fortress: https://youtu.be/MxV24so-dw8

2. How Dwarf Fortress Evolved over 16 Years of Development: https://youtu.be/PiNeDarPzeI

3. How Dwarf Fortress Coming to Steam Changed Everything: https://youtu.be/TX65XCk2-6U anecdote about the money starts @ 18:40

4. Life After the Success of Dwarf Fortress: https://youtu.be/I5zgBXp4e8A

I'm still getting major uncanny valley from any of the videos featuring humans, something about them disgusts me. I guess I should be glad I'm still able to distinguish them.

yeah I don't feel qualified to comment on the study in particular but from personal experience the sort of 'loss of consciousness' does track.

When I realize I've spent an hour doomscrolling in an algorithmic feed, I don't feel like was 'all there' paying attention that whole period, it sort of washes over you and occasionally it rouses your attention and you pop out of unconsciousness to consume a little nugget of 'content' and then dive back in. It feels not worth it to spend so much time 'away' for a negligible amount of entertainment.

Video games or a book feels way different, like I'm also going under, but I'm going somewhere else, to a place that was crafted by another human with aim of it being worth my time.

Yeah me too, and it feels like a lot of people who are exclusively 'stuck' in language seem to think the pre-language part is somehow inferior (which imo is very misguided)

What is the lesswrong-crowd answer to the non-existential risk to working people?

I'm way more worried about these guys breaking the economy in such a way that enriches them and makes me poorer, than I'm about any existential risk these guys cook up in their thought experiments.

Like if you are making more money than god in silicon valley I guess you don't really fear losing your job or house, but for most people that would be quite bad.

I never hear these guys articulate what their concrete vision for the economy if they make ASI / AGI happen (one of those is the correct shibboleth and I don't remember which).

dont worry, without any jobs available the pool of desperate human test subjects will be massive.

As an openrouter user this is better than most alternatives I guess

for someone with a limited budget it is actually very important because it makes me less scared to experiment.

grandma's baking knowledge isn't common sense but expertise.

Do we know what AI startup? I think its important for people to know when they do business with that startup.

yeah for me it's all about specificity, like no one else but Philip K. Dick could have written VALIS, it is so tied to his life and psychology. Its difficult for me to imagine something which is is build on huge amounts of various cultural output to retain the specificity of a singular limited human life-perspective.

I guess the bet is that they will become embedded in basically 8% of the tasks in every sector. which I guess could happen long term? thinking about the whole ass economy in this way always make my mind boggle so I don't really feel equipped to talk intelligently about this.

bad guys doing bad stuff with the model is a worthwhile thing to watch out for, but for me the big question is mostly how these labs are going to make sure that these models will not collapse society. Like mass unemployment + ability to generate incredible amounts of propaganda and disinfo, seems to me like a recipe for incredible amounts of bloodshed. like to me the biggest risks look like Khmer Rouge-ing: "all people who know what matrix multiplication is get the wall"; or something like: Boston Dynamics killbots doing tiananmen square every day.

Of course Americans want us to dismantle the EU, we are even weaker against US influence without it.

strong disagree, as a user the recent updates have been awesome and immediately useful.

Godot is such a nice project, I feel like almost every release brings stuff with it that is directly useful for me. It feels like the main contributors there are really on the ball.

they should try and replace some executives with AI, seems like there is way more room for improvement there.

They probably also have an insane dataset

can you give an example of being defrauded? I don't really know what people mean when they say that.