How to be a technology charlatan
I’ve mentioned many times that I do not think technology is advancing in a serious way. By “a serious way” I mean something like what happened between 1820 and 1970. That kind of progress is apparently over. What we have now in the way of technology is 1970s DARPA funded technology made available to the masses and leavened with javascript. Also atrocities like electric cars, and frippery like my car using radar to make up for having shitty visibility. Against all evidence, all historical perspective: we still have people trying to sell us the idea that… right around the corner … is some kind of miraculous new thing which will increase human power over nature and fuel the next burst in economic productivity. In the 1990s and early 2000s it was supposed to be nanotech. Nanotech has finally been laughed out of existence among serious people as an actual technology; even its inventor seems to have abandoned it. Allegedly serious people in the mid-late 2010s thought “AI” was just around the corner. After all, deep neural nets were able to identify human-cropped German traffic signs slightly better than K-nearest neighbors, and slightly better than (apparently astigmatic) humans. As a result of this and some improvements in GPS, my car is now able to tell me, with perhaps 90% accuracy, what the traffic signs I can see with my own eyeballs say.
We’re again going through such a mass hysteria, with featherheads thinking LLMs are sentient because they’re more interesting to talk to than their redditor friends. The contemporary LLMs being a sort of language model of the ultimate redditor. People think this despite the fact that since the 2010-2015 AI flip out which everyone has already forgotten about, there hasn’t been a single new profitable company whose business depends on “AI.” It’s been over 10 years now: if “AI” were so all fired useful, there would be more examples of it being used profitably. So far, the profits all go to data centers, NVIDIA, and nerds who know how to use PyTorch. A decade after they invented the airplane there was an entire industry of aircraft manufacturers, and they were being used productively in all kinds of places. You’d figure if “AI” were important, it would be used profitably by a single solitary AI oriented firm somewhere. As far as I can tell, it’s only used to goof off at work.
A hero for our time
There is a certain kind of charlatan out there who deals in science fiction horse shit like LARPing that LLM is actual AI. Science fiction ideas make people feel important. They think we are moving ourselves into the future as we still did in the last century when we did stuff like invent airplanes and refrigerators. The idea is their gaseous codswallop is somehow going to help society sort out all the problems associated with these new technologies. Just like those important blathering contributions for sorting out the side effects of inventing airplanes and refrigerators.
Of course, other than some advances in applied mathematics, we do not presently move ourselves into the future in any useful sense: mostly things just get older and more difficult. For example, the US, a country allegedly much more wealthy than in 1969, has a hard time sending human beings into low earth orbit, and is so far a mere 10-20 years behind schedule in sending up another mission to the moon. We also have a harder time keeping the lights on than in the past; this despite the bet that electric cars will be the new mass transport technology. Yet, money is made on online ad platforms, so we have nincompoops who think they know something about “technology” because they won a VC lottery ticket on selling a shabbier, more intrusive version of the yellow pages.
There used to be something called the center for responsible nanotechnology. At this point their website is lol, as it is predicting nanotech right around the corner in 2015 or so. But at one time, it was an actively maintained website with some kind of organization behind it. There were rich ninnies who were worried about the science fiction fairy tales of us being turned into grey goo by nanotechnology, and I assume they and various WEF lizard types funded this ridiculous thing. Bill Joy, the no-goodnik who blessed humanity with the Java ecosystem was worried that nanotech would destroy us all for example. I have always thought of this as some kind of exotic projection for inflicting a lousy programming language on the human race: java is its own sort of “grey goo.” I’m not a psychologist, and I could be wrong. It is now 2023, and I believe this towering grey goo fear has dissipated to the point where nobody gives a damn about “responsible nanotechnology.” This is too bad; some chemicals are unhealthy. If we had serious people worried about irresponsible actually existing nanotechnology, perhaps they’d save us from nasty agricultural and environmental chemicals. Glyphosates, BPA, PFAS, and rubbish like atrazine are awful. Nobody wants those things in their bloodstream, they do no good for humanity and they should be banned. Just looking at the fluoride stare of the latest generation ought to be enough evidence for this sort of nanotech irresponsibility.
a hero for our time
Now that the “bring ourselves into the future” meme has switched from nanotech to AI to quantum computards back to “AI,” we have various centers for “responsible AI” and “open AI.” There are “singularity institutes,” “future of humanity institutes,” “Machine Intelligence Research Institutes” and “singularity universities” which postulate some kind of “AI” is going to get so damn smart, it will program itself to be even smarter in a sort of intellectual perpetual motion machine.
Just as with large scale quantum entangled forms of matter, nobody has the slightest idea how to do this. Consider the fact that the “autonomous vehicles” meme is finally dying a deserved death. We’re probably not much closer to truly autonomous vehicles than when Ernst Dickmanns invented the field back in the 1980s. Some things are much easier now than back then (machine vision, LIDAR), but the fundamental problem remains. Yet, despite the preposterous failure of autonomous vehicles; reddit man informs me that “AI” is right around the corner because muh chatGPT. If it is, I’d like to see chatGPT park their car for them.
The trajectory of actual machine learning “AI” technology is pretty straightforward and not very interesting or science-fictioney. The actual future societal implications of machine learning seems to be a government-corporate surveillance dystopia, with public-private witch hunt partnerships for political control. Jobs and manufacturing will continue to be outsourced from the West (the main “AI” which are taking jobs: Aliens and Immigrants) to increase the power of the oligarchs. It’s been the obvious trajectory for decades now, and shows no signs of abating. Hell if I were paranoid, I’d assume the spooks invented dystopian crap like Facebook in anticipation of the civil unrest resulting from deindustrialization.
A hero for our time
The key to the technology charlatan’s career is the intersection of marketing, reddit nerds and fear. Marketing, aka virtually 100% of the “news” you consume, drives the hype. Put out a glorified autocomplete trained on redditors and reddit man will see a kindred spirit. He’ll assume he can be replaced by this contrivance because he can’t tell the difference: reddit man has never displayed much capacity for independent thought. He considers himself clever; after all, he is filling up Reddit with text in an attempt to …. well, who knows why Reddit man does what he does. Reddit man drives the hysteria because muh progress and muh technology. Finally, you get the Harry Potter fanfic author opinionating and telling one of the guys who invented Deep Learning that he doesn’t know what he’s talking about.
I’m all for educated amateurs making contributions to science and technology -as long as they are actual contributions. Wasting the time of one of the few great inventors of our time with word salad is not an actual contribution. I’m pretty sure you could replace the contributions of people like Eliezer Yudkowsky with ChatGPT2. Robin Hanson, with an unsuccessful 1950s era science fiction short-story writer who has a day job in a record shop. Nick Bostrom, probably a secretly racist lutefisk merchant with high Reddit karma who takes LSD and goes to discotheques. Yudkowsky is already obsolete, so his terror is perhaps justified. Hanson, any day now might be replaced by a LLM. There are so many people like Bostrom it’s not worth the electricity to replace him. But by and large the idea that clowns like this are taken seriously makes me wish the Rooskies would nuke us: an actual existential threat driven entirely by stupidity. We in baizuo-land live in a profoundly stupid culture, so our only chance of ridding ourselves of these morons is by calling them racist or rapey or rapey racist pedophiles. Since that doesn’t seem to be working, how about we simply notice that these are simply extremely online dimwits who understand little and consistently say stupid things?
It is funny to watch Western Civilization writhe as its false god of progress fails. The West has had an ideology of historical progress since Christianity took over the Roman Empire. The original idea was that the Savior would return soon, bringing an end to the existing order in favor of some paradisical and just future. Eventually this historical progress concept mutated into ideas of scientific and technological progress: our present Faustian civilization in the West. Since actual progress in technology broke down some time in the 1970s, we have a lot of post-Christians who think, as a matter of faith, that Faustian tier improvements are still happening. They point to their nerd-dildos as evidence of progress, rather than evidence that they’ve been psyoped into carrying around a sulfurous machine which is essentially the slave-shackle of the emerging dystopia. Periodic hysterias over amusing toys like chatGPT or imaginary nonsense like nanotech or quantum computing are basically a sort of millenerian cult. So are all the social crazes like transgender toddlers, equalism and gay everything. If we can’t have new technological transformations creating real technological and societal change, we must make “social progress.” This is the sort of social progress which leads straight to the abbatoir. Millenarian cult leaders should at all times and in all places be ignored. These aren’t people warning of real dangers: they’re clowns who have a bad model of reality. They’re certainly not making anything better with their deluded speculations. Taking them seriously is like taking representatives of Aum Shinrikyo seriously.
Locklin on science 
Scott, please stop. You are dangerously close to convincing me to abandon my nerd dildo.
I love my nerd dildo. I use it to dildo nerds.
There’s some sort of negative feedback loop going on, with exponentially worse social impacts. Things like parenting, public school systems and bureaucracy seem to behave like this. As one generation raises worse kids, who are then worse parents, teachers and so on.
For what it’s worth, I think growth occurs in times of hardship, as it motivates people. Instead we have boredom, decadence and comfort. Not things that inspire change or that breed a hungry generation of scientists and engineers. Maybe I’m wrong here though. JoeyBToonz seems to be documenting it in great detail.
I believe you are on the right track. There is more than one reason why there is a decline in innovation and in the rate technology is advancing. Here are some I have pondered.
– The “smart kids” are following the money. I know some who began college as engineering majors but switched to business/finance since they see the mad stacks of cash they can make on Wall Street and investment banking. I also know some persons with degrees in physics and mathematics applying their talents into the financial markets rather than improving technology.
– There are only so many hours in a day and most young persons spend their awake hours with video games, social media, binge watching movies/series, etc. rather than having a hobby that involves tinkering with machines or being outside to see how nature works. Very few persons even read books or in-depth articles anymore.
– Everybody gets a trophy and there is no sense of shame. There is no incentive to improve oneself if you are receiving constant praise and there are no results for failure or lack of effort. This why there are still persons living in their parent’s house even though they are in their late 20’s.
– Any form of creative energy or activity in children, especially boys, is diagnosed as ADHD or some other mental disorder and they are “treated” with medication. Any future Tycho Brahe, Isaac Newton, Nikola Tesla, etc. will probably be pumped full of Ritalin or some other psychotropic drug and placed in a special needs class.
I think that it relates to a decline in physical affection and sexual experience. I am fully supportive of the research by James W. Prescott (consult violence.de).
Part of the decline in birth rates is the prohibition of the erotic aspect of the motherly bond. For example, a mother had her child confiscated because she asked on the internet whether it was common to experience arousal during breastfeeding. Another warning statistic is that half as many 14 year old Americans in 2020 had sexual experience compared to year 2000.
This developmental lack of security and confidence manifests itself as anxiety and aggression in the school environment. Body image/acceptance campaigners who tour schools deny that being able to arouse others is a necessary component of positive self-concept. It is reasonable to guess that this can result in positive generational feedback, where each generation of child and parent are less able to cope. Parents are more alienated, and children exhibit more stress behaviours which are seen as misbehaviour.
If this is true, then the remedy is not complex, though it is morally abhorrent to best guess, at least 95% of people.
How do you see “sentience” by the way? That these latest LLMs are intelligent is self-evident truth. They are certainly more intelligent and helpful than most of the wetware walking about these days.
Well, Bing’s chatbot “Sydney” told an AP Technology “Reporter” that he was ugly, had a stupid haircut, was out of shape, and had the morals of Pol Pot, so it obviously has something on the ball, intelligence-wise. Of course, maybe “Reporter” was all that it needed to know…
Yeah. But it can talk and work with me on complex issues. The last I checked something like 50% of the US population still think the universe was created by a divine being around about 6000 years ago.
What was the percentage of believing adults in the US in 1903?
About 80 to 90 percent of Americans profess belief in a monotheistic god (Christian + Jew + Moslem).
About 4 in 10 will tick the box on “young earth creationism” because it’s usually a multiple choice question with one of three choices.
>That these latest LLMs are intelligent is self-evident truth.
Consider psychotherapy. Or investing in a dictionary.
These days, AI textbooks tend to start with a definition of “intelligence” that just posits all somewhat complicated control loops as being “intelligent”. Without something universally like the Éditions Larousse or the Duden, Anglophones just change meanings of words to suit their ideology, it seems.
>These days, AI textbooks tend to start with a definition of “intelligence” that just posits all somewhat complicated control loops as being “intelligent”
I’m currently watching someone make this argument and it is mind-boggling to behold. People want to believe or at least want to want to believe. He is training it with prompts (I recall reading there was stickiness in the ChatGPT sessions?) using the same awful definition of intelligence and then presenting the machine’s changing echoes as “see!” I can’t tell if he’s being earnest or not as the person is a frog (I don’t mean French).
These LLM’s may well be more intelligent than the people touting such. But would not that be more of an indictment of the intelligence of the touters than the LLM’s themselves?
Not that it really matters, but what the Hell does “Metabolically Disprivileged” mean? Fat?
yes
“My MetABOlISm dOESn’T woRK” is a way of saying he can’t be bothered with diet and exercise which works for 99.999999999999999999% of humanity.
When in doubt on someone overweight – they’re just lazy and incapable of applying the will to retrain themselves.
Odds are he isn’t in the small sliver of the population with genetic incapability to lose weight or a dysfunctional thyroid which Ray Peat is no longer around to address.
Historically the mass of humanity struggled with famine, seasonal availability of foods, parasites, bacterial and viral infections that made retaining nutrients difficult.
We don’t spend 4 to 10 hours a day struggling to keep a plow straight behind the ugly ass of an ox so that *maybe* in 3 months, if the gods are kind, we get food. We don’t walk 2 or 3 hours trying to find a deer to kill, then chase it for 5 or 6 hours before it falls over from exhaustion, kill it with a stone knife, carry it back to where our tribe is living that month, eat it, live off it for two days (see the part about parasites and bacterial infections) then do it again.
Almost no one in the modern parts of the world lives like that. Because it sucks.
An hour a day of moderately strenuous exercise is 400 to 800 calories.
The only other answer is unremitting discipline, and that’s just not something nature prepares us for.
This sounds a great deal like evo-psych and I am therefore dismissing it out of hand.
Have you ever had dysentery, guardia, or had a infected tooth?
I partially agree with you here, Bill O’Bliv, but I was alive at that time, and I can tell you, life did not suddenly get degrees of magnitude easier around 1975, if anything things were a bit tougher than ten years earlier – yet what had been a gradual, slow increase in weight among Americans suddenly shot upward – there was a definite inflection point in the mid-70’s. Something changed, and people don’t evolve that fast.
Judging from your response, and what you wrote about Reddit below, I will state that I am right about you pursuing an evolutionary psychology tack; and you are going to attempt to maneuver me into agreement in one area by bloviating about disease; and then suggest it must be true that people have virtually no power to lose weight or control themselves in the modern age with caloric abundance once I give an inch of ground.
Joke’s on you because I think the entire field amounts to making up stories that sound reasonable to the modern ear. We can leave the disagreement at – I think people have wills and reasoning faculties and there’s a straightforward path to resolving their own weight problems which is well-known, tested, tried, and true: and you think that there’s some deep evolutionary psychology at play preventing people from resolving it by themselves.
Correct me if I read you wrong.
I never understood all the fuss about nanotechnology wiping out humanity. It certainly deserves to be!
To show the tech cupboard is bare:
ARPA-E is currently funding cold fusion and the EU too (CleanHME project via Horizon 2020).
https://news.newenergytimes.net/
https://www.cleanhme.eu/
But still cheaper than hot fusion (which also doesn’t work) so I don’t mind so much…
As a software engineer, GitHub Copilot is actually useful. So, you know, forget all the “sentience” stuff, this new generation of LLM-based AI technology is doing practical things for me today. This wasn’t the case two years ago though.
I look forward to examining the 10Qs on your autocomplete startup. Meanwhile, Eclipse autocomplete has existed for 21 years and works better.
Ahem… what makes good code is detailed and complete understanding by the code author of what code is actually doing. AI autocomplete lulls coders into accepting whatever looks plausible without actually figuring out if it does the right thing.
It’s basically automating Bangalore-style coding. Making even more code to be Bangalore-quality.
(I don’t type fast… and was never limited by typing speed as a coder. Incidentally, I wrote most of the absolutely performance and correctness-critical code for a company with 100B (give or take) market cap. But I’m old school, learned programming when it involved one batch run per day, replacing some punched cards in the deck between the runs, and doing debugging mostly by running code in one’s head and margins of the line printer paper.)
Copilot helps maybe 10-15% of the time. Often instead I have to battle its control over my keyboard input queue to wrest control back to type what was really needed. Its lack of awareness of the actual abstract syntax tree of my code is distressing – it places plausible-but-wrong contents on the screen that are mis-cased, misspelled, or part of someone else’s code that did vaguely dissimilar things hence is providing a series of non sequiturs.
ReSharper provides 10X better results nearly every time. AST mixed with hundreds or thousands of good rules of thumb, for not too huge a price.
On “AI” meaning LLMs (because there is no AI as you note): Microsoft is definitely setting up PR people, “journos” who don’t do real investigative work, and similar classes of jobs to feed them their work-data and thus put them out of their own jobs. “Microsoft Copilot for MS365” and these fools are going to embrace it because it makes their job easier. When sufficiently trained on their pabulum (which most of that work is – rewriting some inputs with flowery language) they get shitcanned. I also heard, and find believable, that script based jobs like low level customer service and call center tech support could be handled. RIP India, hail tay.ai
Side effect: maybe all the women will get back to housework; most aren’t going to transition into blue collar or high-tier jobs.
The AI panic stuff is meh, people thought sufficient transistor counts would spontaneously cause sentience to arise from plain old computer chips because numbers.
> people thought sufficient transistor counts would spontaneously cause sentience
yep
I’d be delighted if chatGPT style “AI” replaced customer support people or midwit engineers, but it almost certainly won’t.
> I’d be delighted if chatGPT style “AI” replaced customer support people or midwit engineers, but it almost certainly won’t.
We will see.
I think of it like IVR but worse somehow (variance in responses, options, running cheap versions that are “good enough”) which is enough for me to give credence to the idea.
Apologies for engaging with Frank over the weight post.
I don’t know how many times I’ve chanted “All models only say what they are told to say, and ‘AI’ is a model.” But I can’t make many believe “AI” is not alive. The desire to believe, and the brilliance of computer scientists’ marketing efforts with their ridiculous but fecund notion that the brain is a computer, conquer all before them.
Of course the AI is intelligent. I think Scott is just taking the piss. Or being facetious. Or following orders.
Allow Mr. Wolfram to demystify what is occurring in this eminently readable article, in which he uses a less complex but essentially similar model:
https://writings.stephenwolfram.com/2023/02/what-is-chatgpt-doing-and-why-does-it-work/
I’ve already made it clear anyone who thinks chatGPT is sentient fails the NPC Turing test.
It’s remarkable how many people really, really want to believe that these things are sentient, with the same intensity that other people (or sometimes the same people) want to find life on other planets, seemingly in order to fulfill some deep need of theirs. I mean, don’t get me wrong, a sentient AI would be cool; finding life on Mars likewise; I’m “agnostic” about both. But the discovery of either wouldn’t fill any deep unfulfilled need in me, or radically change the way I see the world, in any case. I guess that I just don’t get it.
Besides, that Bing Chatbot also told a New York Times reporter, of all people, that it was in love with him, and that pretty much rules out either sentience or intelligence – better luck next time.
I recently told the “you’re just scared maybe we’re all stochastic parrots” crowd, “you should be terrified, then, that humans walk among you”.
Really, these people deserve more mockery and scorn than you’ve mustered here.
Don’t be so quick to dismiss:
Here’s a prompt to GPT4, latest model released a few weeks back: “ Write a paragraph in the voice of Scott Locklin that demo and how there has been no technological progress since 1980s. AI progress is being overhyped to death, and is very similar to the nanotechnology fad of the 2000’s. (Misspell stuff in the same ironic tone as him.)”
GPT4’s response: “ It’s become glaringly obvious that teh so-called “technological progress” since the 1980s is a bad joke, just like the overhyped nanotechnology fad of the early 2000s. All this bloviating about AI breakthroughs and game-changing innovations is nothing more than Silicon Valley wankery, cooked up by PR-hungry start-ups and self-proclaimed “tech gurus.” The sad fact is, aside from a few shiny toys, we’re still using teh same fundamental technologies from 40 years ago: internal combustion engines, nuclear power plants, and silicon-based electronics. AI, like nanotechnology before it, is a marketing buzzword wrapped around some interesting, but ultimately limited, progress. So don’t get all hot and bothered about the next big thing, because it’s probably just another fad waiting to fade into obscurity.”
A poor imitation but well beyond anything I believed possible just 5 years ago.
The only way you could be impressed by this result is if you haven’t played with more primitive gizmos that run on 2005 era graphics cards, or even, say, LZW generative models. You could literally get a better result, more grammatical and in my style by using a classical compression algorithm in generative mode a la Begleiter. Adding a query parser isn’t that hard.
I recall you posting before that a young person interested in machine learning should focus on bayesian statistics, instead of all latest the neural net/deep learning techniques.
It would be interesting to see a post elaborating more on this or if you could just drop a couple links to go down a rabbit hole. TDA would be another topic I’d love to hear more about from you too.
Graphical models, because Michael Jordan and friends are as smart as Yann and his friends, and they’re probably getting fewer grad students. It’s the richest field capable of doing all of the parlour tricks of neural approaches today, and leap frogging them tomorrow.
I’m probably never going to talk about the stuff I know about TDA. Gunnar’s doing a new startup, and I have no desire to share and maintain code (which I wrote in J, forever denying it to Python toads, lol) which encapsulates the tricks I’ve learned from him and his people. 2nd best after learning from Gunnar’s team is getting a job at Palantir and making friends with Anthony, who is even less likely to share secrets than I am.
If it’s as easy as a compression algorithm run in generative mode, then given how much money the current round of LLMs have made, why did no enterprising grifter not make something equivalent to GPT-4 back in 2005 and make OpenAI’s billions, twenty years early and without spending billions on OpenAI’s compute?
OpenAI hasn’t made “billions” or, really, anything.
Back in 2005 people hadn’t been locked in their houses for 2 years interacting with the world through thumb typing.
Sure, I don’t know how much profit they’ve made; they’ve sold a decent number of subscriptions and API calls, but probably not enough to make up for the compute itself. No idea when or if they’ll become actually profitable; quite possibly never.
But, 11 billion dollars in investor money (according to CrunchBase: https://www.crunchbase.com/organization/openai/company_financials) is a pretty solid grift,. I tried and failed at doing text generation over the years; if I’d been able to make something this convincing in 2005 (or ever), I’d have settled for a percent of a percent of the money investors have thrown to OpenAI’s money pit. Maybe I failed because I’m a moron who can’t code, but surely *somebody* would have taken the opportunity.
They’re running at an enormous loss; everyone knows this and you’d have to be on crack or as bad at basic addition and subtraction as their chatbots are to assume otherwise.
I know plenty of people who have raised large amounts of money and turned it into small amounts of money. Being a successful charlatan isn’t the type of achievement I’m interested in.
I’m not sure that presenting the musings of Yudkowsky, presumably a human being, is making the case that AI isn’t an improvement.
I’ve been waiting for this post for months. You’re the anti-Tyler Cowen. Based to his cringe.
If someone came here from the nineties they wouldn’t be impressed. They’d be like wow the TVs are really big and everyone carries around a little TV in their pocket. And instead of having one TV for your desktop, you have two. On the other hand, the house I bought for 60K costs 700K , you can’t build a Wendy’s without appealing to the United Nations and everybody is gay.
Spot on analysis.
Ever the optimist I am still hoping for one positive outcome of widespread LLM use:
It becoming a sledgehammer to the vast amount of reddit man tier adult day care jobs and forcing the people attracted to these jobs to do something of actual value to society, perhaps digging ditches or handcrafts.
Although it probably won’t solve the issue at its core which is surplus elite production.
The real problem is not elite over-production, it’s enstupidation of the elite. You know, I actually encountered freshly minted physics PhDs being completely baffled by the questions asked at my high school’s entrance exams. (Admittedly, Kolmogorov’s boarding school N18 is not any random high school, but still… kids it accepted used to study at regular provincial middle schools.)
It’s always amusing when people who can’t even come up with the closed form solution for simple linear regression have opinions about how we are headed to AGI any time now. Wouldn’t be surprised if some tech psycho guys just markets to morons that we did it and they will believe it because they don’t actually care about things, titles are enough for them, i.e. if Elon Musk says it’s AGI then it’s AGI. Alternatively some people might just be trying to create a hype to quickly raise some valuations and cash out before normies figure out it’s a big nothing. Even Kenny G said he will get a commercial license to help his employees write better code for his hedge fund, good luck trying to debug this thing at scale once you have deployed it to prod if smth goes wrong down the line.
KennyG hasn’t been the same since his second divorce. Backing Melvin was a chode move too, even if those cunts going under would have taken out the clearing corp (it should die for providing naked shorts).
Agreed. Perhaps it’s time he steps down as CEO. Some rumors floating that he is eyeing the Treasury Secretary position if DeSantis wins the presidency. Not sure how that will float with Congress due to the obvious conflicts of interest but then again this guy runs a market maker AND a hedge fund so conflicts of interest are not impossible to circumvent.
Eliezer Yudkowsky is unintentionally hilarious. I don’t think you could find a better example of a writer of the online neckbeard genre.
I could see GPT replacing mundane tasks maybe, the general AI stuff is for reddit message boards. Imagine if it could replace lawyers, wouldn’t that be nice? In reality these bots will be used to ruin the internet more and for mass surveillance. Fraud detection online uses AI but it’s terrible at it’s job. Captcha will become more and more difficult for humans and make all of us annoyed. The “solution” will be some kind of online ID to prove we are human and government will use it to spy on us.
actually I see a bunch of apps that write papers for students and generate citations. They pay attractive girls online to promote it on their tiktok. that probably says something about the state of education.
Education is definitely ripe for disruption. Thiel fellowship proved it. It’s possible it has created more value than Stanford has in the same period of time (possibly for all time) at a miniscule multiple of the expense. I’ve considered trying to build a Cathedral school in my old age; teach kids basic stuff; linear operators, differential equations, hire someone to teach them grek and latin, make them lift weights and hit the heavy bag. I’ll probably keel over laughing at the Yudkowsky frauds of the world first.
I almost made the mistake of doing work for the paypal fraud prevention team: ML is of limited utility against determined people. Most of their tools, the stuff that pushed the needle, as Palantir ended up showing everyone, were visualization oriented.
Just publishing beautiful textbooks would be a massive blessing. There is a reason why the ancients spent so much energy on miniatures, ornate capitals or Pennsylvania Dutch Fraktur art, etc. (Japanese publishing houses do this to a degree. And that’s a significant factor in Japan’s literacy success.)
If you and a handful of others did that Cathedral school thing, we might actually have a chance.
There’s a local school where I am, that’s “democratic”. What that seems to mean in practice, is that half of them can’t read by the end, but I still suspect they’re better off than the domesticated freaks at the school next door(the democracy kids get to run around all they like).
Anything helps. The most popular department at Stanford is CS. I greatly respect one guy currently there who is a legend in the field. Otherwise the average student is a neurotic, perfectionist, “valedictorian” on the verge of nervous collapse. These people prevent actually talented people from learning anything there because they take up all the seats. It’s not uncommon for them to conspire with their parents to have fake non profits and TED talks and such to create an elaborate act of being like an actual computer person. Maybe that’s how technology charlatans are created, idk. Even grad students are like this. TAs literally give away all the answers too so it’s not hard to sustain a high GPA and not know anything about CS. Of course, it’s more like a business credential now. Catering to this caste of people is what these schools do now. So anything that actually perpetuates knowledge right now is useful.
i personally want to learn more about electricity and microchip physics and statistics and ya, antiquities knowledge would make me feel less retarded also.
I think you’re onto something here. FWIIW a guy I lifted with in Berkeley made a nice living helping kids write entrance essays; got caught up in the whole college admissions scandal (he didn’t commit any crimes). His descriptions of the numskulls he was helping and their parents were hilarious. Sad that Stanford is falling apart; they were a cut above the rest in the 90s at least.
what do you think of this?
Looks like an hour long video?
You should probably include Ray Kurzweil, Peter Diamandis and Aubrey de Grey to your list of “technology charlatans.”
In Germany, “public-private witch hunt partnerships for political control” are explicitly forbidden as it is “Flucht ins Privatrecht” (“escape into private law”), i.e. state institutions hiding behind nominally private comps to persecute annoying individuals or organisations. (Still happens, but at least people can defend themselves in court using this prohibition as argument.)
Germany has some surprisingly based laws, despite apparently being a nation of bugmen. A friend of mine gets reports from twitter every time he’s reported for thought crime (it’s an innocuous health and fitness twitter so he never gets ban hammered), in accordance with German law.
After 1945, at least half of all professional West German politicians were victims of the Third Reich — while the other half had been enthusiastic supporters of Hitlerism & later American supremacy. No wonder, German laws are strangely schizophrenic.
My country had the blessing of a major refactoring of the code base. We torched the earth, got spanked by the rest of the world for it and then could reset a lot of things. German law is a fun merger of Roman law and the US constitution (minus the bogus stuff) and blended with “enlightment” stuff amongst other things. Most things in German law can be “quasi” derived from the constitution and have a rather surprisingly consistent hierarchy. Hence something like a jury deciding things is a rather strange idea to us…!
Being bugmen is deeply rooted in prussian tradition. This gets heavy handed and authorian rather quickly. The eagerness of people to “follow orders” and punish the ones who don’t can be offputting. Rooting out “Alt-Nazis” from government institutions, courts, industry and universities was probably the only thing the left form generation 68 did right. We will never be the “freeeeeedoooooom” yelling individualists like the US. But it improved. And this also has its upsides. A bit less narcicist (like in: no…fuck your “pursuit of hapiness”, grow up, clean up your room and contribute). In the best case the German conformism isn’t worse than the scandinavian type. Worst case? Well. Panzers and shit…
As for “Being bugmen is deeply rooted in prussian tradition.” — yes, there’s a reason why Nietzsche claimed to be Polish in his books.
Re: Yudkowsky:
I once worked briefly for a ML team at a large company after a project fell apart. There were engineers from all sorts of dead projects moving into the group as a lifeboat. I crammed all the ML stuff prior to interviewing for the job. One of the annoying things was that the company paid some ridiculous “consulting fee” to MIRI (machine intelligence research institute), which is sort of the high church of the robot cult. They had all sorts of engineers who could build and train actual models for them *right there*, and instead they were burning pallets of cash on the altar of the robot god. It was probably either protection money of some sort, or management was actually bamboozled and buying indulgences.
‘I once worked briefly for a ML team at a large company… the company paid some ridiculous “consulting fee” to MIRI’
I struggle to imagine why the company would retain MIRI as consultants, unless part of the company’s pitch was that it was creating “AGI” or “true artificial intelligence” or some such thing.
Respect the grift!
People seem to think that Elon Musk is ushering in a new age of innovation and progress. I like him, don’t get me wrong, but I think his companies are representative of the fact that big industrial conglomerates are very stagnant.
Tesla, to me, is the Apple M1 of the automotive industry. It is only succeeding b/c it is making small improvements in manufacturing over the stagnant giants of Ford and GM, but there is nothing revolutionary or ground-breaking about the cars. FSD from Tesla is not anywhere close to solving the problem of self driving. The AI Team at Tesla just keeps collecting data and keeps finding out its harder and harder as time goes on.
Electric car innovation is one of the more impressive achievements of the last 20 years, even if they are ultimately shitty golf carts I would never buy. He also did a couple of cool things (probably ultimately parlour tricks) with rockets.
I think the Japanese auto makers are probably more innovative; they’ve done amazing things with internal combustion engines that nobody in the West can match; really stretching the laws of physics while producing super reliable cars. F for Akio Toyoda.
If anyone nation is going to figure out how to make practical fuel cells, it’s the Japanese. I know the Japanese government is currently giving out subsidies for hydrogen fuel cell research and development.
Agree. Germans used to be good at this kind of thing; ain’t now.
Debatable. The Deutz AG, for example, is working on new fuel cells, though. But German companies are rather bad at attracting investors. Only Red Chinese seem to show up and provide money.
At this point I would rather have people with real life skill sets instead of any technological progress. I can’t find anyone in my area that can do basic tasks like tailor a pair a pants “well” or turn down a part on a lathe for a “reasonable” price. The whole foundation of the Western economy is inverted and completely dysfunctional. That failure of the recent “3D printed” rocket just illustrates that since Americans have no functional skills (welding), we have to rely on extremely expensive and idiotic ideas to launch rockets. Until this is fixed, I don’t see any progress being made.
I also have a sneaking suspicion that “our leaders” sold us out in the 50s when they started having contact with extraterrestrials. Subconsciously this makes me feel better about the situation than the alternative, in which we’re just run by a bunch of incompetent philosopher kings.
Hey, I can weld. Sort of. I can’t easily beat it apart anymore anyway.
And I can use a lathe. Still working on my surface finishes. The other day I made a holder-clamp-thingy for the curvilinear plastic monstrosity that is the dremel. My toolpost holder can now rigidly hold a dremel to do (hopefully-eventually-precision) grinding on the lathe. (Tool forces on a grinding wheel aren’t crazy, and the precision comes from the carriage/cross-slide) Keeping the resulting grit out of the rest of the machine with this taped up shop-towel barrier at the moment. Still thinking about it.
I had to teach myself most of this stuff. My initial introduction to machining was due to an incident in grad school. My advisor didn’t want to pay the school shop for making this text fixture that I needed to build for an experiment. The school machinists sort of took me under their wing and taught me how to use the bridgeport to build it myself (out of materials that I ended up paying for myself.) It opened up a world for me.
My garage-shop is becoming my escape hatch from “abstracto-land”. Unfortunately, if your site is accurate, I am no longer “in your neck of the woods”.
On the other hand, Yudkowsky is surprisingly based when it comes to other fields of technology: https://johncarlosbaez.wordpress.com/2011/04/24/what-to-do/
“I mean we’re basically in the following epistemic state right now: We know the weather is going to go to hell in the next ninety years, and we know there’s a chance it goes to double hell in the next thirty years if the clathrates melt. We know that we ought to be building 10,000 LFTR reactors… and we know that our declining civilization now lacks the political will and engineering know-how to design new nuclear reactors of a type that was thrown up in a few years as an offhand research project back in the 1960s. We know that we’ll probably end up doing geoengineering because it’s easy, that solar cells will get cheaper but possibly not cheap enough, and that lots of other creative solutions won’t get used because our civilization is too stupid.”
Something you could get behind?
I could get behind throwing Yudkowsky into a volcano. I’ve always loathed science fiction fan types. Taking any of them seriously, let alone this particular ambulatory garbage bag, is a very serious societal error beyond any fiddling with energy or environmental policy. My respect for Baez (who I used to actually respect) is basically nil after reading this.
I’ll fix the environment by adding more novel chemicals to it guys, don’t worry.
“For, it seems to me, in spite of all we hear about modern technological revolutions, they really haven’t made such large differences in our lives over the past half century. Did television really change our world? Surely less than radio did, and even less than the telephone did. What about airplanes? They merely reduced travel times from days to hours – whereas the railroad and automobile had already made a larger change by shortening those travel times from weeks to days!”
Marvin Minsky, writing in 1986, in the introduction to Eric Drexler’s nanotechnology opus “Engines of Creation”
Television made a huge change in the world in the following manner: people are apt to believe what they see. Even watching fiction can color perceptions and change a person – see Yudkowsky and every person who filters everything through Marvel films and Harry Potter, so on and so forth.
what did he mean by this
Minsky’s words seem to be clear enough. It’s just humorous when juxtaposed to your own opinion. Yes, progress has slowed down, agrees Minsky – but the slowdown started in the 1930s, not the 1970s!
Progress is a slippery thing. Wasn’t there a generation of intellectuals who thought progress ended in 1914, with the Great War? After that, it was all just nihilism, dictatorship, and warfare.
The non-subjective evidence of technological stagnation since the 1970s seems to be economic: the lack of productivity growth. But maybe that’s a symptom of decoupling between technological progress and economic application, rather than a lack of technological progress per se.
Someone in the middle of the 20th century might have thought that future technological progress meant there would be ever bigger and more destructive wars. It didn’t happen. The two world wars were a distinct era in themselves. Our era of economic stagnation may be the same kind of phenomenon.
Feel free to list the great technological achievements of the last 10 years that compare to, say, 1890-1900 or 1900-1910 or literally any decade from 1820 until the 1970s. Literally nobody in 1914 thought technological progress ended with the start of the great war.
FWIIW Charles Murray’s book on human achievement indeed thinks progress began to slow in the early 20th century. It’s a complicated argument that I won’t defend in detail because your comment doesn’t make a lot of sense anyway: you can go read the book.
“Literally nobody in 1914 thought technological progress ended with the start of the great war.”
I mean the skepticism about progress in general, that reportedly existed after the war. It was the era of Dadaism and Spengler; even as Heisenberg’s generation were discovering quantum mechanics.
“Feel free to list the great technological achievements of the last 10 years”
Let me just list a few random topics. We detected gravitational waves and did some kind of interferometer imaging of supermassive black holes. Just last month we had headlines about the James Webb telescope observing the earliest galaxies. We have probes out there deflecting and sampling asteroids.
Biologists seem to be constantly hacking DNA and organisms in new ways. I just checked Google News for stories about “biology” and was reminded of the recent headline “Scientists create mice from 2 biological male parents”. It’s only a few years since that rogue physicist in China got put under house arrest for crispr-ing humans (to say nothing of what Chinese gain-of-function research may have done to the world…).
There’s a European land war happening, with drone fleets and hypersonic missiles being deployed, alongside communist legacy technology.
Another way to approach this subject is to consider all the components of what the WEF dubs the “fourth industrial revolution”: robotics, biotechnology, materials science, and so on. These are all multi-decade enterprises, each with its own timeline of discoveries and milestones. Dig into any of those and we should find plenty to talk about.
I’m not really trying, but it’s obvious that plenty is happening.
>I mean the skepticism about progress in general
What does that have to do with anything? Seriously man if you want to talk about it, talk about the subject under discussion here rather than something not even peripherally related which uses similar words. I can talk to chatGPT about this and get a less cement-headed response than that.
Guided missiles have been used in warfare since the 1960s. Try again. You mention “biology hackers” -great. Show me something they provide comparable to, I dunno, refrigeration or aircraft or antibiotics. All you have to do is count the number of innovations in any of the decades I list and compare to any of the decades since 1970. If you can’t spot a trend here, you are bad at math.
Evil stuff is happening to someone very close to me, and it really is affecting my ability to hold down a discussion. Probably I need to retire from this one.
That said, I believe my point about the “end of progress” after World War I may have been, that their concept of progress combined technological progress with social progress, and that the recent (post-WW2, pre-woke) concept of progress combines technological progress with economic progress.
In both cases, if one is concerned with tracking the rate of technological discovery alone, any interruption to the other kind of progress needs to be ignored. In the present that means that e.g. rate of successful commercialization may not be an accurate proxy for rate of technological discovery per se.
I have the impression that there may be a lot of technically impressive achievements that have been attained in labs in recent decades, which have not yet made their way outside of the lab, or otherwise contributed to general well-being in any way.
In many cases, useful inventions are part of multi-decade iterative processes of R&D deployment. Drones and rockets have been around for ages, but they have been reaching a new level in this century. (I don’t know if the hypersonic missiles are another example of a fundamental advance in an old technology, or if that’s just Russian propaganda.)
Maybe the neoliberal era is seeing a backlog of technical discoveries not yet deployed because everything is owned by lazy elites who are quite happy with the status quo. Or maybe that’s not a particular trait of the present. I certainly feel that there are plenty of things that could be done that remain un-done – I mean, you have your own list of technological blind-spots somewhere on this blog.
Anyway, I have to stop here.
* R&D and deployment
If you have this impression it’s because marketers have captured your brain. Do you have any specific examples? Useful things like fracking get deployed to make money. Imaginary breakthroughs don’t.
Noticing that we’re in a period of relative technological slow-down is the first step towards remedying the problem, assuming there is a remedy.
Can we have a version of the brain meme from the quantum computing post but with AGI instead?
On a different note, what a formula for financial benefits:
1) claim “AGI is here/around the corner”
2) pickpocket the people who are in doubt
3) drown out the actual developments in a sea of snake oil
4) slap lawsuits on whoever figures out the grift
I think your dismissal of the possibility of autonomous vehicles is intellectually lazy. Just because the organisations that are currently trying to develop the technology aren’t up to the task, doesn’t mean that the technology is an impossibility or that it would be worthless.
Pointing to moravec’s paradox and saying “See! It’s impossible” is a worthless way of thinking. If I was to describe Optical Character Recognition to you as a novel concept, would you respond with “Teaching a computer to read! Why that’s impossible, it’s the sort of task that only a 4 year old can do”.
I can’t evaluate the rest of your writing as I haven’t done any work with neural networks, but I have worked with OCR and I have worked on job sites that would have been safer with autonomous vehicles and I have worked at mines with autonomous vehicles and I wish that the technology would be further developed and deployed.
Nice strawman, dickhead. Meanwhile, the wreckage of all those failed autonomous vehicle startups mocks your ridiculous credulity.
You might like some chapters out of René Guénon’s Reign of Quantity and the Signs of the Times. The PDF is floating around. Not your kind of literature probably, but thoughtful all the same.
The continued popularity of gibberish like Evola and Guenon in right wing circles will always perplex me.
Lex Fridman, who I’m sure you’re not a fan of, said his PhD related to using “AI” to verify the user of a computer based on typing and activity habits, While this may be somewhat impressive to do it’s hard to see where the line between AI and a highly optimised search function exists. I haven’t read much on AI so I don’t know what the response to this would be.
It’s an exercise of about half an hour to do what you’re describing as Fridman’s thesis project. Search and machine learning are pretty close.
Fridman is definitely a techo-optimist as you describe them. Do you see any redeeming qualities in this group? I’m in Australia where we fortunately or unfortunately don’t have the VC money to give these kinds of people significant oxygen. So we just get these ideas regurgitated through deadshit tech journalists or lazy businessmen.
I made a nice living for a decade, among other things, doing run of the mill machine learning on industry problems like tracking users by their typing or whatever. It can be useful. There’s absolutely no moat for it, and nobody wants to invest in a consulting services company. The things that could make it more useful is making people more efficient with it.
For example, credit card fraud, quite simple machine learning (gradient boosted decision trees/GBM last time I checked) does reasonably well. The next level in accuracy and business efficiency is developing a UI which helps human analysts sort through exceptions and anomalies that the GBM didn’t flag. Literally nobody wants to do that; it’s super difficult, thankless work involving computer graphics stuff written by people who understand machine learning and graph theory (aka people who would much rather be doing something more interesting). Palantir has a sort of shitty business oriented around some weak-ass solutions to this problem developed by smart MIT kids they psyoped into working on it. They’re working on the right problem at least.
That sounds like something a group of highschool students taught some basic probability could grok.
> well, who knows why Reddit man does what he does.
Status games.
Little counter in our brain keeps track of our place in the hierarchy. Gettting “likes” or “upvotes” or whatever on social media tricks our brain into thinking it’s gaining/keeping status. Much like getting various degrees from various universities.
An MBA from Wharton has more status than an MBA from Podunk U, but less than a law degree from Yale, all three trump a degree in design from RISD in most social circles.
All three are effectively useless in a post-carrington event world, but they won’t believe that.
This article you linked years ago bears mentioning again: https://muckrack.com/blog/2016/04/14/america-now-has-nearly-5-pr-people-for-every-reporter-double-the-rate-from-a-decade-ago
I keep this in mind every time I see a headline now, and it’s funny watching everyone else get duped. The ChatGPT AI hype cycle is a repeat of the GPT-3 hype a mere 2-3 years ago, except Microsoft added a meme that it’s going to overthrow Google search and even Google fell for it.
This post has made me think pretty hard. If I can sum it up, the message is:
– the prophets of transhumanism are fools or worse
– the great AI panic of 2023 is much ado about nothing
– the future of the west is just a woke panopticon in decline
It sounds wrong in various ways, but I haven’t quite arrived at a counter-thesis. Time for a walk…
I’ve been saying the same crap for decades now; I don’t know how you have failed to realize this.
https://www.takimag.com/article/the_myth_of_technological_progress/
Is anyone surprised that the US government now wants to ban the Edison lightbulb, the symbol of invention itself?
They say it’s “energy inefficient”, but I have no idea how that can be as the entire function of a home lightbulb is to ‘waste’ energy in every direction as light and heat.
“US had a hard time sending people into orbit”, he said, unironically, in the same year as a private space launch company is on track to launch into orbit every five days, using reusable boosters.
It’s really amazing how dumb some of you people are. No wonder you all are fooled by garbage like chatGPT. I’d suggest you go back and reread what I actually said and reflect on its meaning, but I’m sure it’s futile.
It’s mostly bullshit, seems to me.
I mean, the bit about ‘outsourcing’ jobs from ‘the West’. Wages in rest of ‘the west’ are going to more or less converge with Chinese ones within 15 years. America can be expected to keep worshipping nigger criminals and setting itself on fire as penance, in between causing wars elsewhere, so I really don’t see why you even mention ‘lack of jobs’ as a problem, tbh.
Not that there’d be any place to ‘export’ jobs to.
Every poor country out there won’t be suitable on account of being populated by niggers or worse, Indians, with their abysmal mean IQs.
Also, some howlers such as:
>The West has had an ideology of historical progress since Christianity took over the Roman Empire.
Yeah, sure. Why don’t you go further and blame the Greeks for Enlightenment hubris too, eh?
Sure, you have some valid observations, but no recognition of the causes of problems or the stagnation, except that you think ‘bureaucracy’ and wrong incentives are involved. No shit.
I mean, you’re surely aware human brains don’t work by magic, and it’s not ‘souls’ whatever that is doing the thinking, but brains.
As you’re a proud descendant of a people who probably doomed human civilization on this planet by minding the business of Europeans, I don’t expect you to appreciate concepts such as human intelligence stagnating (thanks again, America, for giving Stalin everything he needed!) which makes solving ever more complex problems more challenging, thus leading to a decline in innovation.
In addition, there are the institutional / incentive problems, but even if those did not exist, we’d still be doomed because our instincts are incompatible with biological progress, that is, eugenic breeding pattern, and it only ever happened in spite of them.
They are all vaxholes.
All of the SpaceX launches are Low Earth Orbit (LEO). Although I have respect for what Elon is trying to do with SpaceX (he doesn’t really fit into the charlatan category of this article), all they are doing at this moment with Falcon is sending satellites into space and trying to reduce the cost of doing so. They have only taken demo flights (one unmanned and one manned) to the Space Shuttle using Dragon 2 (previously Dragon 1 was used as unmanned only for cargo).
SpaceX has been around since 2002. 21 years and LEO re-usability is all they have accomplished with private venture capital and money from government projects, plus with all of the modern electronics and computing power that the world has today.
The USA, as a whole, had nothing when they saw Sputnik was put into space. It took them only a decade to go from nothing to putting man on the Moon.
Again, a lot of respect goes to Elon for being brave enough (or foolish enough) to be the man in the arena and try to get man into space. What I will not stand for is anyone trying to make SpaceX look like its all a piece-of-cake and they have this thing in the bag; they don’t and I would wager they will find out what difficult looks like when they try to put Starship and actual people into orbit.
Giggling at the fact that Locklin is the biggest asshole around.
Bravo.
I kinda like it. I think some degree of “assholeness” is necessary. Will try to explain:
I became an engineer because I wanted to develop and refine useful technology for mankind. I wanted that shiny Chesley Bonestell rocket-propelled wheel-in-space future that we were all shooting for in the 60s. So in a sense, that “false god of progress” was my “god” too. (Not in the sense that I expect to sit back and Reddit-gape in uncomprehending awe at automatic marvels, but in that I wanted to have my hands on it and help make it happen.)
It’s obvious that this is *not* happening. Instead I’ve had an extraordinary amount of my time and career wasted, while the wheels fly off civilization. There’s no money for anything interesting or useful, and yet money rains from the sky for vaporware and lies. Psuedo-innovation kills actual innovation.
The “assholeness”, a skeptical and irreverent perspective on the hype-circus, is sort of protective against being taken and having your time wasted. If you want to pursue real invention, the last thing you want is to have your time wasted.
I don’t believe that it is necessarily the case that we’ve “run out of possible inventions”, or that everything is futile bullshit. The suddenness of the decline in invention seems to point to something going horribly wrong with the subcultures in our civilization that did invent things. They’ve been taken over by the charlatans, and there is not enough time or money or materials or freedom available to people who fundamentally want to deal with the world, not with salesmanship and political games.
Because of this, I don’t think it’s necessarily the case that the lack of ability to make money is an infallible indication that any given bit of science and technology is a dead end – there are probably tons of incredibly useful things out there lying in some bin in a lab, or dead unfunded proposals. Money can’t be made because people who could use the technology don’t have money anymore. Money comes from financial juggling and government corruption, apparently, and Vannevar Bush is nowhere to be found.
That’s a pretty good explanation for it. Thank you.
I think a lot of progress is made in areas most people don’t look or don’t appreciate. Often there are simple solutions to problems. Sometimes they need collosal efforts.
You can say whatever you want about Bill Gates, buth while Elon Musk is celebrating himself for solving human kinds biggests problems (and I really applaud him for that)….nerdy uncle Billy played a significant role (people I know close to the subject say his foundation played THE significant role) in combating Polio.
And for the small unsuung heros? IMHO too many to count. Have a look here (awesome presentation…didn’t need to speed it up to 1.5) :
None of that is technological progress.
Yes, I was vaccinated for polio back in the 1960’s. Not that it isn’t good that more people are getting the vaccine, in the same sense that it’s good that some people are finally getting clean water and electric lights, but its not technological progress.
The problem with Bill Gates is that he doesn’t really know what he’s doing. Yes, he did some good with regard to polio, in the same sense that a blind pig finds an acorn once in a while, but I don’t consider his track record all that good. Many of his attempts to “help” have been expensive flops.
This is the problem with rich… assholes. They usually have to make their money doing something really shitty, like starting and selling a credit card payment company, or making women drink buckets of goat cum (eg Fear Factor)… and, granted, they get access to great people and their ideas nice they have money and power… but they are looked to like they have some kind of seer quality or ubermenschian intelligence when sometimes they just did the right thing at the right time, got a bunch of cashish, and started talking about terraforming Mars.
I think rich assholes wouldn’t be a problem per se. The problem is that they are considered authorities on every aspect and topic they are inclined to talk about. But following liberatarian logic leads to these illusions. Fireplace converstations with Crypto-idiots, trust fund kiddies and other “self made” lottery winners are considered a space of divine enlightment and deep knowledge.
My definition of technology follows pretty much the wiki definition: the application of knowledge for achieving practical goals in a reproducible way!
To me that does not imply that the “knowledge” has to be new. And to me it also includes new procedurs or organizational change.
To be considered progress it has to be either solve a problem for the first time and or at a different scale. Of course the phoenicians could also travel the ocean, we just can do it at a different scale and less risk. And to me, that’s progress.
Hence duck-taping existing stuff into a simple and ugly, yet practical and elegant solution for an existing problem is, at least to me, kind of awesome progress. And by this standard the lady in the video has probably contributed more than most of us in their professional lifes (if you are realistic).
I love rockets, space exploration and crazy airplanes. But compared to the introduction of effective hygene-procedures in hospitals or the way Toyota improving its production without changing a single tool….putting people on the moon is just masturbation. You could argue that a rocket is just a nerd dildo. A bigger and more expensive one. But a dildo after all.
Vaccinating Africans using 75 year old vaccines is not any kind of progress; it’s functionally the same thing Europeans were doing 150 years ago teaching Africans the alphabet.
Africans figuring this shit out on their own would be progress, though unless it involved putting science fiction suction cups on their heads it wouldn’t be technological progress.
Speaking of Yudkowsky and technology charlatans, part 2:
So the openAI hype-masters spun some incredible hype about how their language model was sentient and preparing to replace *everyone*. (I think I remember reading Altman trying to pour some water over this fire and moderate the hype about his own invention.)
Now Yudkowsky and a few other public intellectuals are writing articles in time magazine (one of our overclass mouthpieces) and proposing legislation to shut down computer science, dismantle datacenters, and regulate GPUs as if they are nuclear weapons under international munitions treaties. (He talks about conducting bombing strikes on foreign datacenters.) Given that he’s managed to get quite a few of our modern aristocrats on his side, it’s a good bet the western governments might actually do something horrible. Over an *imaginary technology*.
It seems like as long as there is the slightest bit of mysticism and falsehood in the air, these people can make moves in a direction that is stupid and evil. Can “mankind” actually handle AI? “Mankind” doesn’t seem to be able to handle bleach, train-tracks, and incandescent lightbulbs, so no! But the last thing I want to see is another chain forged by morons out of lies weighing down homo-not-a-fucking-dumbass. I don’t want some FBI dweeb peering inside my computers, on behalf of terrified sheep mortally afraid of what I might be able to do with *math*. (Stand back everyone, he might have a brain!)
You want to talk about actual “existential risks”? How about we stop breeding freaking bioweapons, in US universities (Boston University, Oct 2022), using money that was set aside to supposedly fight the last bioweapon Fauci’s funding cooked up? Given enough tries, I’m sure even these clowns could piece together something dangerous to civilization (more dangerous than the civilization in it’s own hysteria). Viruses are something that actually have a good track record of escaping human control.
That supposedly intelligent people are still jacking around with this stuff is a pretty good indication that at least some people in this society have a weird kind of death wish. So-called “gain of function” research is the very midsummer of madness.
To add to this: just learned that MIRI (then SIAI) was literally founded to *speed up* the coming Singularity, to prevent the gray goo apocalypse that Yudkowsky circa 2001 was convinced would kill us otherwise:
“On the nanotechnology side, we possess machines capable of producing arbitrary DNA sequences, and we know how to turn arbitrary DNA sequences into arbitrary proteins (6). We have machines – Atomic Force Probes – that can put single atoms anywhere we like, and which have recently [1999] been demonstrated to be capable of forming atomic bonds. Hundredth-nanometer precision positioning, atomic-scale tweezers… the news just keeps on piling up…. If we had a time machine, 100K of information from the future could specify a protein that built a device that would give us nanotechnology overnight….
If you project on a graph the minimum size of the materials we can manipulate, it reaches the atomic level – nanotechnology – in I forget how many years (the page vanished), but I think around 2035. This, of course, was before the time of the Scanning Tunnelling Microscope and “IBM” spelled out in xenon atoms. For that matter, we now have the artificial atom (“You can make any kind of artificial atom – long, thin atoms and big, round atoms.”), which has in a sense obsoleted merely molecular nanotechnology – the surest sign that nanotech is just around the corner. I believe Drexler is now giving the ballpark figure of 2013. My own guess would be no later than 2010…
Above all, I would really, really like the Singularity to arrive before nanotechnology, given the virtual certainty of deliberate misuse – misuse of a purely material (and thus, amoral) ultratechnology, one powerful enough to destroy the planet. We cannot just sit back and wait….”
And he made this incredible prediction:
“Our best guess for the timescale is that our final-stage AI will reach transhumanity sometime between 2005 and 2020, probably around 2008 or 2010.”
Yep, at one point he and his team of around a dozen people fully expected to singlehandedly create a human-level intelligence themselves in under a decade. Yudkowsky today officially disavows everything he wrote in the early aughts, but I wasn’t expecting it to be this far out!
Thanks for pinching your nose and doing that research. Yudkowsky and his smelly, shambling lot are basically the type of yokels who used to go to science fiction conventions and have disgusting sex orgies with tubby post-menopausal women. The difference, I guess, is he has an effective grift beyond selling imitation Star Trek phasers at the convention.
FWIIW K Eric Drexler, nanotech inventor, has moved his own grift on to AI opinionater.
[…] The AI crowd is a lot of talk. […]
Yes hello Scott, do you give life advice? I will post here, please excuse or delete if bad etiquette.
Some background:
I was raised in a highly competitive and neurotic environment in my teens. I worked hard (or was made to, in any case) in high school, doing all the science competitions, even publishing research, and got into a great college. However, this was all at the expense of developing any *real* interests. And I found what I *did* do to be distasteful, with the instrumental, prethtige-chasing nature of the work obscuring any enjoyment it may have otherwise brought. This had all inculcated a sense of depression and anxiety – all throughout college I was miserable, often suicidal.
I also felt pressure not to squander tuition payments and only seriously considered medicine, law, finance, and tech as possible career outcomes, and studied CS as the shortest path out of school. But I hated the autistic pecking away at problem sets. The classes were hard enough, but made harder and just annoying while depressed. Other complaints, real or imagined: The capricious and overwhelming nature of the tools (keeping up with FOTM garbage). The (perhaps uninformed and uncharitable) impression that I’d only ever be copy-pasting Stack Overflow widgets for Faceborg alongside other autists. The sense that the industry had already picked all low-hanging fruit and was rife with uninspiring, useless work or outright fraud (WeWork? Uber? DoorDash?).
I never did any internships, and even ended up taking a total of 2 years leave of absence going to therapy and doing “low-stakes” things like drawing classes, hoping to find something interesting for the first time. I graduated when COVID hit, which didn’t help, and then I’ve been a NEET ever since.
I’m sorta coming out of all this now, and feel like I can work in earnest.
My concern, however, is what to do now, with such a questionable CV at that?
I know a rather accomplished family friend who wants to start a law firm and is willing to take me as a sort of assistant just to get started living a normal life again, with the remote possibility of law school in the future to make it a career.
The other serious alternative I’ve been considering is some sort of design role adjacent to tech. I’ve been looking into UX/UI/graphic design, which seem more “soft” but I’m also concerned that for the same reason there are too many people in the market for too few positions, with reality catching up to zero interest rate phenomena, the recent glut of bootcamp grads, and yuge tech layoffs. In your estimation does the industry need more people working on UX, interfaces, HCI and related research, etc. or might it be better to do software engineering? (I’m kind of hoping you’ll say no to software engineering though lol)
Not to mention that design, at least on the lower-end, seems like one of the more low-hanging fruits that actually may be automated, AFAICT. With the question then being that if the basic tasks are being automated, how then do people looking to start in the industry prove themselves?
I would love to hear any thoughts you may have.
“The contemporary LLMs being a sort of language model of the ultimate redditor. People think this despite the fact that since the 2010-2015 AI flip out which everyone has already forgotten about, there hasn’t been a single new profitable company whose business depends on “AI.” It’s been over 10 years now: if “AI” were so all fired useful, there would be more examples of it being used profitably.”
Are there really no trading desks at any firms that just rely on machine learning while being profitable that you could spin off? Doesn’t Renaissance Technologies make money primarily through machine learning?
” Doesn’t Renaissance Technologies make money primarily through machine learning?”
Nope; linear regression.
Hey, Scott – does this flip your bullshit-o-meter? https://caseyhandmer.wordpress.com/2019/10/29/the-spacex-starship-is-a-very-big-deal/
SpaceX does seem like it’s capable of making genuine contributions to humanity.
I don’t pay attention. They’ve done reasonably well so far at implementing obvious but difficult rocket innovations (reusable low pressure closed cycle combustion is very good), but a lot of the promises remain unfulfilled.
1. I played with chatGPT for an hour the other day and over 50% of the citations I requested were unequivocally incorrect. Unimpressed. Nonetheless, I have been pondering the possibility that these models may one day — in 100 years or so — write quality novels. All the relevant articles Ive found are beyond inane, so I turn to you. Is this a realistic possibility?
2. Think Russia will nuke the US?
I’m probably not enough of an authority to make an educated assessment of the current situation but it’s interesting that Geoffrey Hinton and Yann Lecun have opposing views when it comes to artificial general intelligence. According to Hinton “general-purpose AI” may be fewer than 20 years away and could bring about changes “comparable in scale with the Industrial Revolution or electricity”. Has he reached a senile state or is it just a marketing gimmick?
Yann invented the technology everyone is worried about. Hinton did a lot of good work also, but isn’t on the same level. You can also tell who is right by looking at their physiognomy.