@self_made_human's banner p

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users  
joined 2022 September 05 05:31:00 UTC

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


				

User ID: 454

self_made_human

amaratvaṃ prāpnuhi, athavā yatamāno mṛtyum āpnuhi

16 followers   follows 0 users   joined 2022 September 05 05:31:00 UTC

					

I'm a transhumanist doctor. In a better world, I wouldn't need to add that as a qualifier to plain old "doctor". It would be taken as granted for someone in the profession of saving lives.

At any rate, I intend to live forever or die trying. See you at Heat Death!

Friends:

A friend to everyone is a friend to no one.


					

User ID: 454

Nobody knows how this going to play out.

I've been on the AI x-risk train long before it was cool, or at least a mainstream interest. Can't say for sure when I first stumbled upon LessWrong, but I presume my teenage love for hard scifi would have ensured I stumbled upon a haven for nerds worrying about what was then incredibly speculative science fiction.

God, that must have been in the early 2010s? I don't even remember what I thought at the time. I recall being more worried about the unemployment than the extinction bit, and I may or may not have come full circle.

Circa 2015, while I was in med school, I was deeply concerned about both x-risk and shorter-term automation induced unemployment. At that point, I was thinking this would be a problem for me, personally, in 10-30 years.

I remember, in 2018, arguing with a surgeon who didn't believe in self-driving cars coming to fruition in the near future. He was wrong about that, Waymo is safer than the average human driver per mile. You can order one through an app, if you live in the right city.

I was wrong too, claiming we'd see demos of fully robotic surgery in 5 years. I even offered to bet on it, not that I had any money. Well, it's looking closer to another 5 now. At least I have some money.

I thought I had time to build a career. Marry. Have kids. Become a respected doctor, get some savings and investments in place before the jobs started to go in earnest.

My timelines, in 2017, were about 10-20 years till I was obsolete, but I was wrong in many regards. I expected that higher cognitive tasks would be the last to go. I didn't expect AIs scoring 99th percentile (GPT-4 on release) on the USMLE, or doing graduate level maths, while we don't have affordable multifunction consumer robots.

I thought the Uber drivers, the truckers, they'd be the first to fall under the wheels or rollers of the behemoth coming over the horizon. I'd never have predicted that artists would the first to get bent over.

If your job can be entirely conducted with a computer and an email address, you're so fucking screwed. In the meantime, bricklayers are whistling away with no immediate end in sight.

I liked medicine. Or at least it appealed to me more than the alternatives. If I was more courageous, I might have gone into CS. I expected an unusual degree of job security, due to regulatory hurdles if literally nothing else.

I wanted psychiatry. Much of it can be readily automated. Any of the AI companies who really wanted it could whip up a 3D photorealistic AI avatar and pipe in a webcam. You could get someone far less educated or trained to do the boring physical stuff. I'd automate myself out of 90% of my current job if I had a computer that wasn't locked down by IT. For a more senior psych, they could easily offload the paperwork which is 50% of their workload.

Am I lucky that my natural desires and career goals gave me an unusual degree of safety from job losses? Hell yes. But I'm hardly actually safe. One day, maybe soon, someone will do the maths to prove that the robots can prescribe better than we can, and then get to work on breaking down the barriers that prevent that from happening.

I'm also rather unlucky. Oh, there are far worse places to be, I'm probably in the global 95th percentile for job security. Still, I'm an Indian citizen, on a visa that is predicated on my provision of a vital service in short supply. I don't have much money, and am unlikely to make enough to retire on without working several decades.

I'm the kind of person any Western government would consider an acceptable sacrifice when compared to actual citizens. They'd be right in doing so, what can I ask for, when I'm economically obsolete, except charity?

Go back to India? Where the base of the economy is agriculture and services? When GPT-4o in voice mode can kick most call center employees to the curb? Where the average Wipro or TCS code monkey adds nothing to Claude 3.7? This could happen Today AD, people just haven't gotten the memo. Oh boy.

I've got a contract for 3 years as a trainee. I'm safe for now. I can guess at what the world will look like then, but I have little confidence in my economic utility on the free market when that comes.

somewhat esteemed rationalist?

Eh? Who is esteeming him? Man's getting like 10 likes on average for his Twitter posts.

This controversy is his most popular post by a country mile.

If "somewhat esteemed rationalist" is a term that can be handed out this generously, I'm going to start calling myself that.

I'm skeptical of this as an explanation in this instance, if only because of the fact that if he was predisposed to believe in such nonsensical ideas (whether due to autism or anything else) I don't see how he'd ever have gotten to the state of being taken seriously as a rationalist in the first place. After all, the topic of men vs women in sports won't be only contentious issue he's come across where there's a strong social incentive to take one side over the other.

I was not familiar with this gentleman, and I'd call myself unusually steeped in rat culture, at least as close as you can get without knowing any in person. He's written a few minor SF anthologies, and has an unremarkable number of Twitter followers.

He's not a rationalist big name, nor even a major one. It's not like we hand out ID cards, anyone can call themselves a rationalist.

Despite being called a "disciple of Yud", the only evidence for that claim are is a blurb in his blog:

Recently, his biggest influences have been Paolo Bacigalupi, Peter Watts, and Eliezer Yudkowsky. They all produce rather “thinky” fiction, and the first two are very grim. He has a soft spot for the darker, grimmer side of science fiction.

Is he a disciple of Watts? Am I, because I've read both him and Yudkowsky?

He's nobody. Not even on LessWrong. At best, he "promoted" a popular Medium post there, which I don't think he wrote.

Read the article Memetic Tribes and Culture War 2.0 which Eneasz promoted on LessWrong. Read the Diary of an Aspiring Writer, by N J Magas. So this non-story is that some autistic random person has a silly belief. What of it?

That and running a HPMOR podcast that probably has like 10 listeners is all the real influence he has in the rationalist community.

I know no end of non-autistic people who have ridiculous beliefs. I have an uncle who's a fan of homeopathy despite being an acclaimed microbiologist. He's perfectly capable of switching off Science Mode when he leaves the lab and takes his sugar pills.

One swallow doesn't a summer make, and this is more of a chicken painted blue.

Also, I know you're the psychiatrist, but wouldn't being autistic make it less likely you'd have the requisite cognitive machinery in place necessary to delude yourself about the state of the world for the purposes of social signalling?

In general, yes. That doesn't mean that autistic people are infallible arbiters of truth. Especially social truths/lies, where they're perpetually perplexed by how people just seem to make shit up, when they realize the discrepancy.

At the very least, he wasn't deluding himself for social signaling. He's a victim of those who do, guilty of taking them at face value and lacking enough contact with the physical world to overcome that. He's not getting any status from this disclosure, quite the opposite, everyone is coming to laugh and sneer at him.

I've seen people say that just saying "Ghibli style" works way more often than when you append the Studio before it.

Altman had initially said they'd roll out to all users. But he recently clarified that it's been delayed due to extreme demand from paying customers. Us free-loaders need to sit tight a while longer.

I'm glad someone else is taking on the LLM-posting mantle.

I'd previously discussed Gemini 2.0 Flash's native image gen. It was a hit on Xitter, but nothing close to the tsunami of Ghiblification we're witnessing. Having (paying) normies use your stuff on an app makes a difference.

For a lot of people, it's a massive step-up over what they were used to (DALLE-3 via ChatGPT).

From eye-balling generations, 4o is better than Gemini in almost every aspect. More aesthetic, better prompt adherence and less censored (!!!). That's despite Google having taken a far lighter touch when it came to content moderation.

Human artists NGMI. They had a few years to seethe and cope, while holding up the failures of AI image gen such as fingers or text. It's an effectively solved problem now, well past the minimum viable product into something that casual users are immediately jumping on.

Product photography? Stock photos? Solved. Commercial photography and art is in for a bad time.

It's worth noting OpenAI has been sitting on this for 9 months. The first demos of native image gen were a good while back, and it's obvious that the release was perfectly timed to steal Google's thunder.

I expect a tit-for-tat, with the various companies trying to escalate. This is great for the consumer or informed user, but my condolences on if you rely on this to make a living. It's coming for me too.

100%. Vaping isn't something I'd recommend people start, but when switching from a cigarette habit, you're doing yourself an enormous favor.

My handwriting? Average for a normal person. Good by doctor standards! It's usually clear and legible, if not ornate, but you can tell that by the end of a long day and work it's devolving into something you can interpret as an ECG.

It's probably more consistent now than it ever was when I was trying to actually "write" it versus "drawing" it, if that distinction makes sense. I rather like it, personally. It's satisfying to quickly swoosh it out on a restaurant bill or the like.

Sounds appealing. I'm fond of the aesthetic of whipping out a fountain pen and just going for it haha.

There's a very good reason the country is so popular for sex tourism. It's safe, and it's cheap, and the only English required is me love you long time.

Even if outright brothels are illegal, there are entire red light districts, and you can just take them to a hotel. But of course, if you really want legible prices, legality with easy access and English speakers, your options are limited. I just think those are unusual priorities and people seem to do fine there.

My 5 star hotel that I was staying at with my family (!) overlooked the red light district in that part of Thailand. Trust me, not speaking the local language was the last thing dissuading the gents who came by, and they seemed to be having a great time.

Surely there must be a cheaper way for you to pay to get laid, wherever you are? If you're traveling to Australia at significant expense, there's likely a whole bunch of cheap Asian countries along the way.

You, high IQ, well educated, traveled and read motte denizen, you personally?

Uh.. Yes? I mean, is this a question worth asking really?

I mean, I wasn't born believing in HBD, I was won over. I used to think that terraforming Mars was a great idea, and now I think it's a rather suboptimal choice when it comes to establishing robust space colonies. I used to think that AI would more inhuman, and very much didn't expect them to speak like us before they thought as well as they do.

I'm sure there are plenty of cases where I've been wrong and thus changed my mind. I don't think I've had any drastic collapses of my cognitive framework that forced me to re-evaluate everything.

And the long essay where Eneasz Brodski at request of his readers and haters explains how it happened

This gentleman is autistic. I'm sure we have more than our fair share, but that's a condition that predisposes a tendency to take things at face value without considering how much of it is virtue-signalling or social fiction.

"No, it cannot happen to me! I was trained in martial arts of rationalism by ancient master Yud the Yumongous! I am unstoppable!"

I think you're engaging in the hobby of making up people to be mad at. There are worse hobbies, I'm sure. I'm quite certain that there isn't anyone here who will claim identification with this, unless someone spins up an alt. If there are, I offer my psychiatric services, first interview free.

At any rate, I find singling out Rats and rat-adjacents like the Motte's users as examples of bad epistemics or miscalibration is somewhere between laughable and preaching to the choir. Name a group more obsessed with evaluating the rigor of their beliefs about the world. If someone listened to Yud or Scott and came away with the belief that they themselves were therefore unimpeachable, then they can read an IKEA manual and assemble a mouse-trap that takes their finger off.

If you think we're bad, have you seen the rest of the internet?

Anyone hate their signature, or is it just me?

Well, hate is a strong word, but I see people penning all kinds of fancy scrawls, while mine is just initials in cursive with a small flourish. Even the other variant on some legal documents is an initial and surname in the same.

This can be convenient. God knows most professions don't need to sign their names nearly as often as medics do.

I suppose I could change it, if I really wanted to. I'm just put off by the mild logistical headaches I might face if someone asks me why existing samples on various documents don't match what I put elsewhere.

Scott has a recent post up about it.

https://www.astralcodexten.com/p/the-ozempocalypse-is-nigh

It doesn't take much effort or knowledge to reconstitute it with bacteriostatic water or inject it. I presume you could take it orally too, but I'm not 100% sure about that.

It seems to be way cheaper, and probably safe unless you look at extremely shady sources. I'd do it, if I felt like I needed ozempic, but I also have the option of visiting India and getting it at about $100 for a month's supply.

I don't think even heavy use of my vape appreciably diminished my appetite. I certainly didn't lose any weight because of it. Thinking about the people I know who smoke or vape, they're more likely to be obese than not.

If you're that keen on losing weight, then you can source Chinese semaglutide as dried peptides for the same expense and save yourself the addiction.

I started vaping, on a rainy night when I was stuck outside my Airbnb in London. My ex had cancer, so we spent the wait reading through the literature and finally convinced ourselves that the harms were minimal, compared to cigarettes at the very least. Putting anything but virgin air in your lungs has some downsides.

That being said, I mildly regret developing a nicotine dependence. When I was forced to stop cold-turkey, it was best described as a male version of PMSing or roid-rage.

The upsides? It gives you something to do when hanging out with smokers. It gives me a small boost when I run away to the toilets for a quick puff at work. Nicotine buzzes are fun.

On the balance, I think it's better not to start. I think I read too much Gwern.

My examples like brushing teeth or wearing clothes were intended as a reductio ad absurdum against the knee jerk reaction that any modification or augmentation of the "natural" human state is inherently suspect or falls under some grand, unified theory of "trans*ism". If we define transhumanism simply as "using technology to overcome a limitation," then yes, by that absurdly broad definition, a crow with a wire hook is a transcrowist, and anyone who wears shoes is a transhumanist. That wasn't my serious definition, but rather a way to point out that the boundary between "natural human" and "technologically augmented human" is not some bright, sacred line. Humans have always used tools and modified their environment and even their bodies.

If you are able to note that entirely normal behaviors that most humans, and even some smarter non-human animals engage in can be distinguished from transhumanism despite lying on a clear continuum, then please consider the obvious differences between it and transgenderism.

After all, I'm not the one who asked:

Tell me again how transgenderism is a totally different thing from transhumanism

If you can't appreciate the rather minimal and nuanced differences between getting a haircut and uploading a brain into a computer, then I suppose that's a reasonable question to ask.

There's no bright line between "fixing issues" and "making things better than they were, and on average, are". If you went to an ophthalmologist for LASIK when you developed myopia, you wouldn't get a refund if your complaint was that he'd ended up giving you better than 20/20 vision.

The fact that I don't go about saying how desperately I wish to be the opposite sex and how I could achieve that with modern technology makes me rather different from the average trans ideologue.

That seems to be the major similarity between between transhumanists and transgenderists

There you go. Majority similarities aren't the same as equivalence, and implies major differences. Communism isn't Fascism even if they're both authoritarian in practice. White ethnonationalism isn't the Nation of Islam.

The majority of trans advocates believe that the changes we can make at present, through makeup, surgery, hormones and asking other people nicely, are enough to justify treating a transperson just like a natal individual of the same identified sex. This has next to nothing to do with transhumanism.

Good, 9 inch heels and a urgent medical care when I trip and break something is expensive enough as is!

You're just wasting my time, and have been for a while.

I've explained my stance, that moral views are just as inherently subjective as opinions on favorite colors. That's it.

I was wondering whether to discuss it, but felt too lazy given that it was 1 am. Like you, I'd rather wait for R2, I always try and use reasoning models for anything complex, barring trivial answers

It turned out I must have already watched that video as YouTube shows that I liked it haha.

Of course, everything I've said is speculative, but it's modestly informed speculation. All future advances are sci-fi until they're not, we'll have to strap in for the ride and see where it takes us.

Consciousness isn't computation - it's fundamentally embedded into the biological processes. It also doesn't emerge from neural networks regardless of how well they mimic behaviours of real humans.

Hang on. Please note that you're just saying these things.

Why isn't consciousness computational? I mean, I can't prove that it is either, but that's just assuming the opposite side. The correct stance is agnosticism, albeit I think it'll turn out to have a mechanistic explanation eventually.

How do you know with, any confidence at all, that something like an LLM isn't conscious? It might not be conscious in the same manner as humans, but the same might be true for birds or octopi. They demonstrate all the hallmarks of intelligence, even if it's not the human kind.

Neural networks are statistical models, while you are your un-statistical emotions, you are your hormonal systems, microbiomes, other physical systems within your body.

The human body operates on biology. Which is abstracted chemistry. Which is abstracted physics. Which can be mathematically modeled. There's nothing in the human body or brain that violates the laws of physics, you need supernovae or massive particle collides to produce behavior the Standard Model can't explain (leaving aside dark matter and energy, which aren't relevant to human biology).

That physics, while intractable to compute at the quantum level or even the microscopic scale, still holds true. A lot of it can be usefully described and wrangled with statistics.

If you just extract the consciousness + the memories, just the raw contents of your brain and put them into the machine, you lose everything else, which is arguably the most important part.

Well, emulate the body too! The neurons are electro-chemical, and surprisingly binary, in the sense that they're either firing or they aren't. This behavior can be well approximated.

If a disconnected brain emulation goes nuts, then if you have that kind of tech, you can trivially design a virtual body with the usual sensory modalities.

The brain is also very noisy. You can probably get away with saving a lot of computation by approximating events at the chemical scale. Not every random jiggle of proteins matters, it simply can't at scale.

All of this irreducible complexity can't be reimplemented by assuming that everything is an algorithm.

I am not convinced that this complexity is irreducible at all. A single neuron, or even a thousand, misfiring: Happens all the time. Doesn't matter. An emulation can withstand a lot of noise, because the object it's representing is also noisy.

Emotions aren't algorithmic abstract patterns, they are complex interactions between neurons and other biological systems and they are a fundamental part of the biological reality that makes you "you".

Emotions can be both algorithmic patterns and the product of a complex interplay between systems. All that really changes is that the algorithms in question become more complex.

This is already accounted for. Estimates for the amount of compute needed for a brain emulation vary multiple orders of magnitude. I never said this would be easy. It just isn't impossible, look, something evolution cobbled up manages. We even have our own alien artifical intelligences that can run on your phone.

Omitting them makes you a spider, an alien.

If that was really the outcome, I'd take it over death when my fragile biological form fails me. I don't think this is likely at all, beyond the first imperfect uploads.

Thanks. I didn't realize they'd finally released 4o image capabilities.

When it comes to text generation models, I can't bring myself to care anymore. They all do the basics proficiently, which is generate code snippets and console commands and trivia answers to questions that I can ask in a few sentences. And they all sort of fail on really complex stuff like doing my work for me in making changes to a 1M line codebase.

LLMs have been "good enough" for my use cases for a while now. I haven't seen anything like the jump from GPT 3.5 to 4, where prompts that previously had been borderline useless magically became useful. Not that I mind incremental improvement, I'd say that reasoning models were a decent change in terms of QOL and reliability when it came to complex reasoning.

If I had to use an LLM for my day job, then pretty much any recent model would work fine. While medical records can be massive, dense and a PITA to work with, they're not as complex as a large programming project. The whole field was built around humans needing to keep the gist of things in working memory. I'm more constrained by things like data protection laws, access points idiot-proofed by IT, and the lack of tooling to make it easy to transfer records or information back out. Ah, the wonders of the NHS..

I remain quite confident that LLMs will continue to improve, and suspect a lot of the recent sense of being underwhelmed that I've experienced is because the rapid rate of iteration makes changes less stark. They're beating agentic behavior into them, more by the minute.

Another day, another LLM

Google just launched the latest iteration of their Gemini language models. I suppose the name was most appropriate for the (short) period where the version number was precisely 2.0.

Well, everyone say hi to Gemini 2.5 Pro (Thinking). The naming scheme hasn't gotten any better, albeit Google beats the packs's abysmal average performance. It's a narrow win, folks, as there's a Gemini 2.0 Pro, a smaller and leaner 2.0 Flash, and a 2.0 Flash Lite.

(We're days away from OAI matching them with an o3-mini-high-low-too-slow.)

What stands out about this model? Nothing really. It reasons by default, which can be nice, but at the cost of increased latency for responses.

It is incrementally better on benchmarks, but even Google's PR team couldn't drum up a revolutionary new capability to showcase. They get a pass, because 2.0 Flash's image gen was revolutionary, and happened a mere week or so ago.

Gemini models have recently become the Honda Civic of LLMs. Not nearly as flashy, but reliable and with no obvious downsides. This one has seized the number one spot on LM Arena's leaderboard, based off (nominally) blinded user feedback. It might hold on to it for a week, or a month. The days when GPT-4 retained the crown for months on end are gone.

After plenty of use, all I can confidently say is that it writes better. I'm very happy with that. I'm sure someone will find a task it can do better than the rest, but I doubt it'll make anyone switch over if they're already happy. I'm confident there's something deep to be said about my inability to meaningfully differentiate models in terms of capability, be it for work or play. I'm just not going to be the one to say it today.

I don't want my kids to have the option to do heroin, even if it's paired with a pamphlet explaining the real likely outcomes.

What if they need surgery and a helpful anesthetist asks them if they'd like some diamorphine for the pain and hands them a brochure? That's heroin.

However, I don't even think that that's a viable option. Seems like our options are: ban and demonize heroin, or legalize it and subsidize its use (as was recently done in British Columbia).

I like neither option, or at least not the "subsidize its use". We live in a world where heroin can be legally used, in the appropriate context. I reject all the bad options and advocate for the good ones. If that doesn't suffice, then I'll choose the least bad one, as I presume you would too.

Same with transitioning kids: I don't see how we ever get to a world where it's both legal and the pros and cons are presented honestly. So I think I'd rather throw the few kids who could conceivably benefit from it under the bus and ban it for everybody.

There are plenty of psychiatric interventions I grapple with at work that require careful scrutiny and vetting! Electroshock therapy was highly controversial and demonized. It's still being used as ECT, and saves lives.

A prospective filter would have to be very expensive to outweigh my other goal, which is to increase personal liberty. Someone who doesn't value the latter is entirely correct if they have a lower threshold for giving up early. We would have the issue of having similar values, but weighting them separately depending on the context.