keltan's Shortform

I would bet hard cash that "I’m sure I’ll be at least in the bottom 5% for intelligence at Less Online. I won’t be surprised or hurt if I’ve got the least Gs of people there" is wrong.  

But okay, let's just go with your fears and assume for sake of argument that you're right...  Last year,  Aella did a live polling event at Manifest where she asked us to, in general, first predict how we relatively compared to the crowd on any N, then define actual buckets of values and arrange ourselves by those buckets.  

(So, e.g., "predict how recently have you exercised compared to others; the most recent, all the way on the left, the least recent, all the way on the right" and then "okay, let's define from that fence to that chair as 'in the last hour', the chair to the steps as 'in the last day', etc." and have folks line up twice based on first their relative guess, then their actual value.)

I was overall deeply surprised when she had folks line up under that system by SAT scores and IQ scores.  It was, generally, extremely poorly correlated with how awesome it was to get to talk with a given person.

This post already expresses a great deal of the vibes that make me think it will be awesome to meet you, and that you get a lot of the vibes folks are aiming for, as far as I can tell.  Looking forward to meeting ya!

Re safety, I don't know about Oakland but some parts of SF are genuinely the most dangerous feeling places I've ever been to after dark (because normally I wouldn't go somewhere, but SF feels very fine until it isn't). If I am travelling to places in SF after dark I'll check how dodgy the street entrances are. 

Have fun! I won't be going. Some random notes:

• berkeley, san francisco is like saying maitland, newcastle
• re: #1: yeah I don't feel unsafe outside after dark in the bay. If a homeless person walks by, I'll just say hi and ask if they have any urgent unmet needs. even just acknowledging them as a person is a nice gesture, though. many will try to engage much more than you have time or interest for; it's okay to just walk away from the convo.
• had to look up what "getting rolled" is. yeah, it's possible, but not that hard to avoid. if an area seems very poor, there will be more desperate people. but the highest risk of being robbed is probably opportunistically on the train. keep your eyes mobile; it's probably a 1 in 300 to 1 in 3,000 train trips event, but it's pretty annoying when it happens, to put it mildly.
• I'm not aware of there being an intense presence of organized aggressive groups in oakland, but there's certainly plenty of disorganized aggression, again mostly from desperate people. I got out of what was going to be a mugging once by offering to send them internet money (venmo) before they asked for anything, and they were so knocked off balance by this (I was saying "I don't have cash but I can send it on an app") that they almost bolted instead of accepting it. carry cash if you want to share it on purpose (people ask for money a lot and it feels nicer to say yes than no); don't if you don't. it's not as bad as some places though, because the warmth means less desperation from homeless folks; homeless folks are usually pretty chill, if rather upset at the system. there is a specific ongoing aggressive presence: there are organized car-breakin and bike theft rings. but I don't think it's like gangs you may have heard about in the past in LA. the theft rings generally want to grab the thing and get the fuck away, not engage. if you hadn't asked and nobody had told you, you probably wouldn't even have noticed anything besides harmless homeless people mumbling something they think is interesting under their breath and not expecting to be understood because they get ignored by everyone.
• yeah homeless people often have tents. it's not where a civ would hope to be, but tents are just houses. treat it similarly.
• is it a cult: you tell me whether it has the bad patterns that define cults. I'd personally say there have been cults spawned by it, but it's more of a general community, with reasonably healthy community patterns. Don't (ever) let your guard down about cults, though, in any context.
• you should be worried someone convinces you to move to the bay. it's not worth it. like, literally entirely for cost of housing reasons, no other reason, everything else is great, there's a reason people are there anyway. but phew, the niceness comes with a honkin price tag. and no, living in a 10ft by 10ft room to get vaguely normal sounding rent is not a good idea, even though it's possible.
• average bay area people are definitely overachievers, see above about cost of housing. this is not true of america in general.
• the most important california warnings are about weed: don't buy weed. DON'T USE INHALED WEED. edibles can be a bad time if you take more than you think you're taking, but won't ruin your whole life as long as you go in with steadfast rules about when you have them, and rules like not buying them yourself. in fact, never use an inhaled or injected recreational drug, period - the fast uptake is extremely dangerous and will likely actually knock your motivation system off balance hard enough to probably ruin your life. you probably won't be offered weed unless you ask for it, and even then most people won't have any to share. If they do, it might be because they have a bad habit. It's a fun drug when contained to a social setting, though. if someone has some I might suggest trying 2mg or less (ie, one fifth chunk of a normal 10mg edible), even if you're used to weed it's not the vibe I'd suggest for highly technical conversations.

I think go ahead and ask people to sign things. I've done it before and it went great, and the resulting book is a great memento. You've got a good conversation starter right there with asking them their favourite sequences post.

Welcome to the US!

Re: safety, it depends on exactly where you are, your skill in assessing strangers' intentions from a distance, and probably the way you carry yourself.

Speaking of which, I'd be interested in playing some improv games with you at less.online, if you want to do that!

How did you like your trip in the end?

Welcome to the US; excited for your time at LessOnline (and maybe Manifest too?)

And re: 19., we're working on it!^[1]

1. ^{^}

   (Sorry, that was a lie too.)

Are there gangs walking around Oakland looking to stab people?

Please don't ask on Manifold, you might incentivize creating one.

Have you tried whiteboarding-related techniques?

I think that suddenly starting to using written media (even journals), in an environment without much or any guidance, is like pressing too hard on the gas; you're gaining incredible power and going from zero to one on things faster than you ever have before. 

Depending on their environment and what they're interested in starting out, some people might learn (or be shown) how to steer quickly, whereas others might accumulate/scaffold really lopsided optimization power and crash and burn (e.g. getting involved in tons of stuff at once that upon reflection was way too much for someone just starting out).

Maybe I could even write a sequence on this?

Which was not terribly secret. The details of the Project were indeed super-secret, to the point where most of the politicians hadn't known anything, but despite the massive global-scale censorship & secrecy, many had observed the signs of a major project of some sort and some got as far as a nuclear bomb specifically. Also, commercial satellites with meter resolution did not exist which could quantify major facilities or new cities like Los Alamos or Hanford (but overflights, and then satellites, now exist and have helped reveal later top-secret nuclear bomb programs). An AI Manhattan Project, as we currently think of it, would be amusingly similar in footprint (eg. energy consumption) to the original and often observable from space: all those gigawatts have to go somewhere, after all.* I'm sure you can find plenty more about secrecy breaches in Rhodes.

This was not necessarily all that useful in the context of WWII - of course America had some big secret projects going, everyone did. It was a total world war. Everyone was aware there was a war on. The devil was in the details of what the program was - a failure like the V2-s, or a success like Enigma decrypts and Manhattan? But a binary exists/does-not-exist is useful in a peacetime context and the current discussion.

(If nothing else, the fact that DeepSeek keeps publishing is a signal. I would note here BTW that you cannot argue, without tying yourself into some pretzel knots explaining 4-D chess logic, that Chinese AI is about to catch up to and surpass the West because the best Chinese AI group, DeepSeek, just released a model or published this-or-that revealing the secrets of OA, and argue that there is already a secret all-out Chinese Manhattan Project going on which will potentially reach AGI first - because the first thing the latter would have done is stop the former from publishing anything which might help Western AI and then devour it for researchers.)

* A wag on Twitter has pointed out that the total energy/heat output of something like a GPT-4 or GPT-5 training run is the same as or larger than the output of a Hiroshima/Nagasaki-scale nuclear bomb explosion. Which is helpful intuition for why your datacenters need so much cooling, at least.

Before the internet?

I think the idea is that there would be clues, particularly talented engineers not doing public work. Of course, I don't know who's carefully tracking that for Chinese scientists.

Scientists did guess the Manhattan project existed because several top physicists had ceased publishing. I don't know if that made it to the relevant governments. That would be vastly easier to do with the internet - if anyone is bothering.

Good luck getting the voice model to parrot a basic meth recipe!

This is not particularly useful, plenty of voice models will happily parrot absolutely anything. The important part is not letting your phrase get out; there's work out there on designs for protocols for how to exchange sentences in a way that guarantees no leakage even if someone overhears.

Hmm.  I don't doubt that targeted voice-mimicking scams exist (or will soon).  I don't think memorable, reused passwords are likely to work well enough to foil them.  Between forgetting (on the sender or receiver end), claimed ignorance ("Mom,  I'm in jail and really need money, and I'm freaking out!  No, I don't remember what we said the password would be"), and general social hurdles ("that's a weird thing to want"), I don't think it'll catch on.

Instead, I'd look to context-dependent auth (looking for more confidence when the ask is scammer-adjacent), challenge-response (remember our summer in Fiji?), 2FA (let me call the court to provide the bail), or just much more context (5 minutes of casual conversation with a friend or relative is likely hard to really fake, even if the voice is close).

But really, I recommend security mindset and understanding of authorization levels, even if authentication isn't the main worry.  Most friends, even close ones, shouldn't be allowed to ask you to mail $500 in gift cards to a random address, even if they prove they are really themselves.

I haven't tried harmful outputs, but FWIW I've tried getting it to sing a few times and found that pretty difficult.

In my experience, most of the general public will verbally agree that AI X-risk is a big deal, but then go about their day (cause reasonably, they have no power). There's no obvious social role/action to do in response to that.

For climate, people understand that they should recycle, not keep the water running, and if there's a way to donate to clean the ocean on a Mr. Beast video, then some will even donate (sadly, none of these are very effective for solving the climate problem though! Gotta avoid that for our case).

Having a clear call-to-action seems relevant. For example, educating the public about AI taking jobs for the purpose of building support for UBI. It's then clear what to communicate and the call-to-action.

I'd be curious to hear what you think an ask should be? 

Alternatively, you could argue that generally informing folks on a wide scale about the risks involved will then allow general public to do what they believe is locally best. This could involve a documentary or realistic movie.

What do you mean with "meaningful action" regarding climate in the 2010s?

If you look at solar energy, panel prices go down largely in a straight (on a logarithmic scale) since 1975 with a short pause between 2005 to 2010. German pro-solar policy started in the 1990 and the biggest change in 2011 was the new five-year plan of th CCP which was probably driven more by economic justifications.

There are few people who currently think we should do whatever is possible to reduce temperature rise. Doing so would mean to have a plan for geoengineering. Climate activist often use climate as a justification to push for anticapitalist policies that they independently believe.

The fight for human flourishing doesn't end at the initiation of takeoff [echo many points from Seth Herd [LW · GW] here]. More generally, it's very possible to win the fight and lose the war, and a broader base of people who are invested in AI issues will improve the situation.

(I also don't think this is an accurate simplification of the climate movement or its successes/failures. But that's tangential to the point I'd like to make.)

Sounds like that would encourage theft.

I'm curious why you opted for Aristotle (albeit "modern") as the prompt pre-load? Most of those responses seem not directly tethered to Aristotelian concepts/books or even what he directly posits as being the most important skills and faculties of human cognition. For example, cold reading, I don't recall anything of the sort anywhere in any Aristotle I've read. 

While we're not sure Aristotle himself designed the layout of the corpus, we do know that in the Nicomachean Ethics lists the faculties of "whereby the soul attains Truth":

Techne (Τεχνε) - which refers to conventional ways of achieving goals, i.e. without deliberation 
Episteme (Επιστήμε) - which is apodeiktike or the faculty of arguing from proofs
Phronesis (Φρονέσις) - confusingly translated as "practical wisdom" this refers to the ability to deliberate to attain goals by means of deliberation. Excellence in phronesis is translated by the latinate word 'Prudence'.
Sofia (Σοφια) - often translated as 'wisdom' - Aristotle calls this the investigation of causes.
Nous (Νους ) - which refers to the archai - or the 'first principles'

According to Diogenes Laertius, the corpus (at least as it has come to us) divides into the practical books and the theoretical - the practical itself would be subdivided between the books on Techne (say Rhetoric and Poetics), and Phronesis (Ethics and Politics), the theoretical is then covered in works like the Metaphysics (which is probably not even a cohesive book, but a hodge-podge), Categories etc. etc. 

This would appear to me to be a better guide for the timeless education in Aristotelian tradition and how we should guide a modern adaptation.

I asked for further details on the 10th point and Claude listed a bunch of stuff I’ve absolutely never heard of. I’d say it’s probably related to meditation if I had to guess. Here’s that.

—

Claude Let me break down Time Perception Management into its deeper components, as this is one of the most subtle yet powerful micro-skills...

1. Understanding Psychological Time vs Clock Time

• How different mental states alter time perception
  • Flow states can make hours feel like minutes
  • Anxiety can make minutes feel like hours
• Learning to intentionally shift between time perspectives
  • When to use event-based time vs clock-based time
  • How to enter "timeless" states for creative work
• Reading your internal clock
  • Understanding your natural rhythms
  • Recognizing when time perception is distorted by emotions

2. Managing Time Horizons

• Developing multiple concurrent time frames
  • Thinking in day/month/year/decade simultaneously
  • Balancing immediate tasks with long-term vision
• Time horizon switching
  • When to zoom out to years
  • When to zoom in to minutes
  • How to maintain strategic clarity while handling tactics
• Future-self continuity
  • Techniques for making future consequences feel real
  • Building strong mental connections to your future self

3. Chronotype Optimization

• Understanding your natural energy patterns
  • Peak creative hours
  • Peak analytical hours
  • Recovery periods
• Designing your life around your chronotype
  • When to schedule different types of work
  • How to handle obligations that fight your natural rhythm
• Seasonal adjustments
  • Adapting to changing daylight
  • Modifying routines for different seasons

4. Timeline Compression and Expansion Techniques

• How to make time feel longer when needed
  • Breaking experiences into smaller chunks
  • Increasing novelty to slow perceived time
  • Using attention to expand moments
• How to make time feel shorter when beneficial
  • Chunking similar tasks
  • Creating flow states
  • Using routine to speed through necessary tasks

5. Personal Tempo Awareness

• Understanding your natural pace
  • Speaking tempo
  • Walking pace
  • Decision-making speed
• Tempo matching and mismatching
  • When to match others' tempo
  • When to intentionally mismatch
  • How to set the tempo in group situations
• Rhythm management
  • Creating sustainable work rhythms
  • Building in recovery cycles
  • Recognizing when to speed up or slow down

6. Advanced Applications:

• Using time perception in negotiations
  • Creating or relieving time pressure
  • Understanding how others perceive time
• Memory engineering
  • Making good times feel longer in memory
  • Making difficult times feel shorter in memory
• Social time management
  • Managing group time perception
  • Creating shared time experiences
• Decision timing
  • Understanding when to decide quickly vs slowly
  • Recognizing optimal timing windows

This skill set is particularly valuable because most people are unconscious of how their time perception affects their decisions and experiences. Mastering it gives you significant advantages in:

• Personal productivity
• Emotional regulation
• Strategic planning
• Social influence
• Life satisfaction

Would you like me to elaborate on any particular aspect? Or shall we explore how this connects with some of the other micro-skills we discussed?​​​​​​​​​​​​​​​​

Without a detailed Model Card for 4o it is impossible to know " for sure" why models drift in performance over time, but drift they do.

It is entirely possible that Open-AI started with a version of GPT-4 Turbo, parallelize processing and performed an extensive "fine tune" to improve the multi-modal capabilities.

Essentially, the model could "forget" how to complete prompuppies. Workhfrom just a week ago, because some of its "memory" was over-written with instructions to complete requests for multi-modal replies.

I'm confused by what you mean that GPT-4o is bad? In my experience it has been stronger than plain GPT-4, especially at more complex stuff. I do physics research and it's the first model that can actually improve the computational efficiency of parts of my code that implement physical models. It has also become more useful for discussing my research, in the sense that it dives deeper into specialized topics, while the previous GPT-4 would just respond in a very handwavy way.