aFiercePancake
No bio...
User ID: 4111
aFiercePancake
1mo ago
STOP DOING THOUGHT EXPERIMENTS
- FAKE THINGS WERE NOT SUPPOSED TO BE DISCUSSED SERIOUSLY
- MILLENIA OF HYPOTHETICALS yet NO REAL WORLD USE FOUND for discussing made up scenarios
- Wanted to talk about something that hasn’t happened for a laugh? We had a tool for that. It was called GUESSING.
- “Yes please give me A WORLDWIDE VOTE USING A SET OF BUTTONS THAT HAVE THE POWER TO INSTANTLY KILL PEOPLE BASED ON WHAT BUTTON OTHER PEOPLE PRESSED.” “Please give me A TRAIN TRACK THAT FORKS IN TWO, SOMEHOW IS UNABLE TO STOP, FOR SOME REASON FIVE PEOPLE ARE TIED TO THE TRACKS AND ONE PERSON TO ANOTHER ONE, AND FOR SOME REASON YOU HAVE A LEVER TO CHANGE THE TRACK THE TRAIN IS ON BUT CANT JUST UNTIE THE PEOPLE” - Statements dreamed up by the utterly deranged.
LOOK at what Philosophers have been demanding your respect for all this time, with all the books and charts we built for them.
This is REAL philosophy done by REAL philosophers (and internet wannabe philosophers)
Chinese_room.jpg
Violinist.jpg
Veil_of_ignorance.jpg
“Hello I would like to talk about what I think I would do in an entirely impossible imaginary scenario”
They have played us for absolute fools
aFiercePancake
3mo ago
Math professor Daniel Litt discusses the challenge here.
Summarizing, these are unpublished lemmas from math professor’s own work. Lemmas tend to be minor theorems or helper results used as little pieces to help prove larger more interesting ideas. So at least somewhat novel (supposedly), but not grand theorems or crucial results. Litt says that in his field, figuring out what lemmas to prove is the hard part, and then proving them is typically much easier. He said that overall proving results like these take up a relatively small fraction of his time, but a tool to automate their proofs would be very helpful.
As for the problems themselves, Litt said they vary greatly in difficulty. Two of them, including two of the 5-6 that Aletheia got right, apparently had nearly identical statements already proven in the literature. Another one had the proof sketched out in literature, but no model managed to fill in the details.
What most interests me is reliability. Litt writes that overall a lot of garbage was produced, and in a ‘real’ scenario where no one actually knows the answer that is a serious problem. Both Aletheia variants didn’t answer 4 of the questions, either because the model said “I don’t know how to solve it” or it hadn’t finished in the allotted time. I couldn’t find the breakdown from a quick skim, which is a shame - I would be very impressed if the model said it couldn’t solve it rather than giving a wrong proof. Still, it seems of the 12 solutions submitted by the two variants to 6 problems, 3 were considered substantially incorrect, for a ‘precision’ of 75% and a ‘recall’ of 45%, given the problems that it didn’t give an attempt for.
Overall, I would say this is better than I would have expected (not that I have any particular insight to the problems themselves), but still seems like it will pose an immense difficulty when these tools are applied to actual problems where the solution isn’t known ahead of time.
aFiercePancake
4mo ago
Sorry if you addressed this elsewhere in the thread, but I’m curious your thoughts on birth rates and its long term impact.
If you tried to reduce a whole country into a single metric of greatness, I think you can argue over whether China or the US is higher, but I’m sympathetic to the argument that the derivative of this metric is clearly greater for China. What about the second or third derivative though? The biggest thing that makes me think the US’s might be higher is birth rates. According to the UN as summarized by Wikipedia, the US is substantially below replacement at 1.6, but China is at 1.0, which seems catastrophic to me. Indeed, there’s news article from this week saying China now has more deaths than births (a cursory search shows this is not the case for the US after a bit of die-off of elderly people during the peak of the pandemic). If these numbers continue to hold (or decline at the same rate) then 2 generations from now the American cohort will be larger than the Chinese one. My sense, though this could be Western propaganda, is that the US is also more attractive to and welcoming of immigrants than China (averaged over a few presidential administrations anyway) and the US has relatively high fertility religious subgroups that are still compatible with modernity like Mormons and some Jewish sects (does China have any equivalents? I honestly don’t know. I suppose the Uyghurs might have been).
The result is that I could see China overtaking the US on a number of metrics in the next couple decades, but then slinking into economic and cultural decline as the younger Chinese generations become smaller in relative and absolute terms compared to the US. Some on this board may believe that will be irrelevant due to AI and robotics, but I’m skeptical of that and certainly wouldn’t bet the farm on it if I were in the Politburo.
But I also grant I have a very siloed information diet on the topic - is there something I’m missing? Maybe population size doesn’t matter that much? After all, during the US’s period of unquestioned dominance it had a much smaller population than China. Still, I think the birth rates point at something important, and China is definitely doing worse in that regard at least.
aFiercePancake
5mo ago
Almost embarrassed that this is what pulled me out of lurkerdom, but when duty calls…
First, if you want to play Skyrim without actually playing Skyrim, try Enderal. It’s a free mod that changes it into an entirely new game with improved mechanics and good worldbuilding. I enjoyed it maybe more than Skyrim itself. There’s a few areas that you can tell were left underdeveloped, but it’s overall polished and has few bugs (for me at least).
For actual Skyrim, I last played a few years ago. Bethesda has continued to update the game by incorporating Creations into the official release, so no idea if these still work. Here’s my top few:
- Ordinator is a perk overhaul that makes things much more interesting than “expert level spells use less mana.”
- SkyrimSouls makes it so that opening a menu doesn’t pause the game (you can configure which menus). Prevents the classic cheese wheel spam (or at least forces you to bind it to a hotkey and make combat more reactive)
- Apocalypse and Summermyst add a bunch of new interesting spells and enchantments, respectively.
- SkyUI is the UI that should have shipped with the PC version, rather than just porting over a console-friendly UI
- Genuinely Intelligent Soul Trap is a small but massively nice QoL improvement.
- Inigo is a fully voiced companion that is fun to travel with.
There’s a burgeoning scene of AI NPC chat in Skyrim: speech to text or free-entry type your dialogue, it goes to an LLM with some prompt engineering to answer as the NPCs, then a text to speech model have them respond back. Obviously a lot of jank involved and there can still be latency issues, but it looks very cool. Mantella was the first big one, CHIM seems to be the most developed currently, and SkyrimNet is a promising up-and-comer. I haven’t actually used any, so do report back if you try them out.
My guess is their LLM over indexes on the recent search history and what you click on. So likely WhiningCoil vaguely described the incident with perhaps incorrect info and with the low information query Google returned bad results. He clicked on them to see if they were the thing he was thinking of and the LLM got that irrelevant stuff stuck in its context. I have similar issues when using OpenAI models professionally and personally.
More options
Context Copy link