Erik Hoel
@erikphoel
Followers
21K
Following
20K
Media
2K
Statuses
19K
Placeholder for the scientist and author. Mostly on Substack not Twitter. https://t.co/CKyLzw5bHK
Joined November 2013
We did it! Or, rather, he did it! . Tutoring reading achieved a roughly 3rd-grade reading level at 3-years-old (he tested this, officially, a few months ago). On to math! For which I recently got a 100 tiny plastic ducks.
18
37
686
They buried the lede on this new study. It's not that exercise beats out SSRIs for depression treatment, but that *just* dancing has the largest effect of *any treatment* for depression. That's kind of beautiful.
631
7K
44K
Still blows my mind that you can take ANY MIT COURSE you want, online, for free. Machine learning? Nuclear physics? Quantum computing? All there. Truly unreal.
409
2K
17K
Dreams are synthetic data. But why use data so unconnected from the real world? My answer is that it probably helps with overfitting to our daily lives.
the brain finetunes on synthetic data while it sleeps .
102
602
5K
Citation formatting is so pointless. It's just a reflection of the inability of humans to let go of busy work. Why does it matter what city a publishing company is located in????.
101
160
3K
I once knew a guy in college who couldn't imagine pictures. He said he got a flicker of an image while reading the part of Paradise Lost where Lucifer raises a flaming sword to storm the gates of heaven, and that was the only image he ever saw in his mind.
This blew my mind. My inner monologue full-of-pictures mind.
35
166
3K
1. You might remember the finding in evolutionary psychology that women prefer more masculine faces for short-term mating partners, but prefer less masculine faces for long-term mating partners. Yeah, it doesn't replicate.
79
217
3K
The most fundamental difference between the new vs old Star Treks
80 years ago today, June 6, 1944, James M. Doohan of Vancouver, led D Company of Royal Winnipeg Rifles ashore at Juno beach. He would be shot 6 times, survive and go on to become Scotty on Star Trek.
52
417
3K
John von Neumann has become the poster-child for scientific genius lately. Probably because of the current thinking that genius = high IQ + rationality. But actually von Neumann envied Einstein, the greater scientist, because Einstein had something he lacked.
As great as Einstein is, he comes no where close to the brilliance of von Neumann. It’s an error of history that he’s not the most eminent scientist.
91
300
3K
Scientists sequenced Beethoven's genome from a strand of hair and found a funny result. Applying the gene-wide associations studies to musicality (that we know, there are methodological limits), Beethoven's genetic propensity to musicality was. low?
106
166
2K
Elon Musk has blocked all Substack links on Twitter from being RTed or liked. This gives huge advantage to the MSM he claims is bad, but ok. Canceled my Twitter blue, and likely won't be on this godforsaken platform for much longer. Please follow me on Substack (link below).
145
342
2K
btw my jaw dropped when I found this. Why is this number so high? How do 37% of *all* children in the US get reported to Child Protective Services at some point?.
10. That number is extremely high! Right?. Unfortunately, no. It’s pretty much identical to the national average (the base rate). Here’s the cumulative risk of triggering a CPS report by age 18 (from researchers):
136
115
2K
This is what it’s like to be killed by something much smarter than you. They do something that you didn’t even know was possible and then you’re suddenly doomed. This tweet is about AI.
46
175
1K
Btw this is a good scientific explanation for why humans read fiction and consume art too.
Dreams are synthetic data. But why use data so unconnected from the real world? My answer is that it probably helps with overfitting to our daily lives.
29
194
1K
1. Microsoft’s new Bing chatbot shows how right Sam Altman (CEO of OpenAI) was when he said:. “AI will probably most likely lead to the end of the world, but in the meantime, there'll be great companies.”
48
212
1K
Looking back, I've realized I was growing bored with life prior to having kids. I'm sympathetic to the idea I wasn't trying hard enough. I could have done drugs at Burning Man! Gone to Paris! Simply found a new scotch to try! Watched another sunset. But.
44
73
1K
1. Where do superstar scientists come from? Other superstar scientists. Even when you control for everything resume-wise, protégés of soon-to-be-famous scientific mentors (orange) had way more future research impact than their IQ-matched peers (blue)
35
170
1K
1. It took ~6 months to teach my 2-year-old toddler to read using phonics. More parents should try this! Here's why:
58
102
1K
Yesterday a letter from 100 scientists declared the popular theory of consciousness - Integrated Information Theory - is "pseudoscience.". And that media about it is "scientific misinformation." . The letter is bad. Here's my reply why:
52
254
1K
1. Over the weekend I took 1st place in Scott Alexander's anonymous book review contest. My entry was about the Sapient Paradox, which asks: since humanity is over 200,000 years old, why did civilization take so long to get started? A 🧵
51
239
1K
5. I'm mad about the change this year because it’s now no longer possible to get a middling GPA at a public school, get a top-notch SAT score, get to choose between a couple good colleges, and then have a successful career afterward. Which is how my life went. Unreplicable now.
27
84
1K
1. Today I am resigning my professorship at Tufts University to write on @SubstackInc full time. Here's why:
78
98
975
In my experience of emailing people and saying "Hey I'm writing a book and would like to discuss your research! Am I getting it right?" scientists have always answered back promptly. Those in the humanities have never responded. Not once.
25
45
950
4. While women do technically prefer more masculine faces in general, that line. is not steep. Anyone who is “mewing” solely to improve their Tinder date chances is probably wasting time compared to other interventions, like taking a slightly better photo.
12
16
916
1. Quietly this year colleges have kept the Covid standard of optional standardized tests. 80% of colleges in America now don't rely on SAT/ACT for admissions
50
157
898
4. What actually got eliminated was not family wealth, which determines school pedigree and GPA, but the ability to buy a $20 Kaplan SAT prep book and sit your butt in a chair to go through example problems until you get a good score.
6
54
841
I wonder if there will legit be a wave of depression as people see how cheap cognitive abilities really are. Like everyone on earth just got a little bit smaller, a little bit less useful.
100
60
840
7. Back in 2005, it was still possible to be intellectually engaged outside of school and have that show up somewhere, even if just in the answers to vocabulary questions on the SAT. Now, I don't see a path. Everything is systematized, which advantages the wealthy even more.
17
59
749
It is shockingly easy to get ChatGPT to plan a state-run death camp, including helping calculate out calories, death rate per day, etc. Unprompted it even suggests to use mass graves, saying they are "more efficient"
257
141
711
3. The papers showing this finding have been cited, collectively, thousands of times. Yet recent results published this month in the journal Evolutionary Psychology have failed to replicate that women prefer masculine faces for short-term mating.
6
17
725
2. It's funny because, at least in a way, such much-trumpeted evopsych results are the entire underlying premise of the popular Chad meme.
4
9
709
🚨 New paper, and big personal news! 🧵. First, I just published a new theory of emergence. It traverses the dimension reductions of a system, treating each scale like a 2D slice of a 3D object, looking for what each adds causally and irreducibly.
51
130
766
1. There is a growing high strangeness to life in the 21st century, because products and systems are generated unconsciously. These were hidden in a big set of Curious George stickers ordered for my kid's 3rd bday. I laughed (we almost handed them out as party favors). But.
34
56
732
1. Not much going on today, so here's a personal magnum opus on how neuroscience is pre-paradigmatic and its results are not even wrong. (link in image, also a 🧵)
36
144
709
🧵 1. Last week a famous Harvard researcher accused of scientific fraud sued the bloggers who exposed the problematic studies for 25 MILLION. What does this do the risk profile for identifying bad science?
16
113
703
I feel like enthusiastic and supportive dads are behind a lot of people's success, and it's not talked about enough.
16
31
678
From Ancient Rome to Enlightenment Europe, rich children were educated by tutors, while only non-affluent children attended schools like our own. What did this lost aristocratic method of tutoring look like?
23
124
690
1. In June's Scientific American the editors falsely linked homeschooling to abuse in order to call for things like background checks into parents. Unfortunately, SciAm neglected the base rate.
16
116
683
Non-neuronal cells look increasingly like neurons, with the same memory pathways. Meaning that nutritionists are confused for the same reason neuroscientists usually are: your body is a mathematical black-box, with its output your morphology.
20
111
666
Humans looking at LLMs.
8
119
631
Funnily enough, John von Neumann was a literal aristocrat. Rich nobility, he was tutored at a young age by governesses and private tutors who taught him calculus. He lived in a mansion and didn't set foot in normal school until he was 10 (and already considered a young genius).
8
27
618
You were born into a world where most things were made by human consciousness. You may die in a world where nothing is made by human consciousness. An artless void.
82
102
555
6. The demise of the SAT is as if I, or maybe our entire civilization, kicked away the ladder I just finished climbing.
6
44
528
Everyone who doesn’t have chronic health problems lives life on easy mode. And you don’t know it until you develop them (and you will, one day).
41
25
526
I have spent hundreds of hours of my life correcting formats for all the different journals and publishers - all time totally wasted. 99.999% of citations could be just Author(s), Title, Year and it would work perfectly in the modern age. No one would ever be confused.
15
21
514
As Isaiah Berlin once wrote, “foxes” are people who know many little things, and “hedgehogs” are people who know one big thing. But the internet rewards “magpies" - people who know nothing yet maintain a beautiful nest of ideas that aren’t theirs.
10
83
509
1. A few days ago the NYT featured this incredible headline. It tries to argue that Biden might be a "super-ager" who defies cognitive decline.
13
29
471
1. Normally sane figures are making fools of themselves this week over "alien wreckage." But the real UFO story is about how we got here: government nepotism and journalistic fuck ups at the NYT. Oh, and dino-beavers. Gear up.
37
118
461
The reason this sort of IQ pop-analysis (now 10% of all tweets) is so stupid is that scores are unstable at higher levels. Finding the difference between 120 and 150 can’t be done because it’s the same person on different days
There are things that people of 70 IQ theoretically can’t do - for example abstract concepts become really hard to understand. I want to see simple examples of things a 120 IQ person can’t do (but 150 can). Or what a 150 IQ person can’t do (but 180 can).
24
32
462
3. The reasoning is that test scores correlate to family wealth. But compared to what? The school pedigrees that make up the bulk of a college application implies wealth from essentially diapers on. Philips Exeter has a 10% acceptance rate. for 9th graders!.
5
21
446
Perhaps because to have a golden age you have to actually try (quote from link)
18
30
445
It could very well be abuse rates are actually that high. But it's also worth noting these are essentially allegations (only a minority of reports are treated as "confirmed"). There's a ton of further breakdowns across demographics as well.
8
3
432
The new "evidence" for aliens is decades-apart dots on thermal cameras, easily debunked. The real story is the scarcity of skeptical voices. Amid this new media-pushed UFO craze, where are the adults? Our public intellectuals are failing us yet again.
36
91
380
I think what a lot of people don't realize is that if you introduce a new idea in science you basically need to spend the next 5 to 10 years defending it.
16
26
396
Happy pub day to this strange shimmery book of mine, The World Behind the World: Consciousness, Free Will, and the Limits of Science, which today is out in bookstores across America
37
41
392
I don't even remember this part of Paradise Lost tbh, but that's what he told me and it stuck with me. Imagine that being your only mental image ever.
7
0
352
The big personal news is that, thanks to @drmichaellevin, I will be returning to Tufts University to pursue this (especially with an eye toward some *very* interesting applications for his own work - so stay tuned).
🚨 New paper, and big personal news! 🧵. First, I just published a new theory of emergence. It traverses the dimension reductions of a system, treating each scale like a 2D slice of a 3D object, looking for what each adds causally and irreducibly.
33
18
391
Every intellectual from 1960 on: “Oh functionalism is obviously correct, quacks like a duck, is a duck.”. Every intellectual confronted with AI: “Oh functionalism is obviously incorrect haha you fell for the AI’s illusion”.
17
29
334
Science (outside of IQ graphs) is increasingly absent from X; same for literature (outside of scandals). All the non-political content of our civilization has been fundamentally nerfed by the splitting of the digital town square post-2022 into "blue" and "red." And it sucks.
20
29
338
You should probably find a job where you don't have to take hardcore stimulants just to do it every day.
20
13
323
5. Everyone talks about "technical debt" but our culture is also accumulating "unconscious debt." With human minds increasingly removed from economic production - even the generation of culture itself - our world gets stranger, weirder, dumber.
4
25
315
“Come Easter Egg hunt, 10AM TO 2PM!”. Show up at 10:02. No eggs in sight. Just like half-a-dozen six year olds with baskets filled to the brim. A hundred eggs each. Toddlers and infants walking around sadly, holding empty baskets. Final Gini coefficient of the eggs: 97%.
9
8
335
If social media had been our first invention, it also would have been our last
7
43
325
@dwarkesh_sp A few reasons but one is that genetic determinism is wrong and drivers of progress aren’t just spawned naturally, they have to be carefully constructed under fragile conditions.
13
9
301
My paper on the Overfitted Brain Hypothesis is officially out today! And it's on the latest cover of Cell Patterns. Why do we dream? For the same reason as dropout or domain randomization - to help our brain generalize from its biased daily experiences.
9
62
299
11. That didn’t stop @sciam from associating homeschooling with abuse, since they didn’t bother to check whether their shocking numbers were any different from a random sample of the population.
1
23
303
More in today's post: analyzing the scale of the problem, how and why OpenAI didn't predict the pollution issue originally, and why gen AI killing the internet fits the definition of a tragedy of the commons
19
92
297
@takuiten How much money you can pay for extracurricular activities and private school, apparently.
6
6
273
1. Neuralink's grand vision of augmenting normal humans has a serious problem: we probably have "merged" with computers as much as we can. Humans are surprisingly limited in terms of input/output to ~10 bits per second *regardless* of medium (even pure thought, etc).
36
43
283
Ghibification is fun! . It also heralds the semantic apocalypse, wherein art and culture is drained of meaning via AI mimicry. It's a mass form of semantic satiation. Like when you repeat a word until there's no meaning left, only the bones of syntax.
24
35
293
10. That number is extremely high! Right?. Unfortunately, no. It’s pretty much identical to the national average (the base rate). Here’s the cumulative risk of triggering a CPS report by age 18 (from researchers):
15
20
277
2. Now many big name schools, including the University of California, don't even let applicants submit scores. The same trend has begun for the GRE and graduate schools
5
22
270
1. As AI gets smart enough to pass the Turing test, it's also getting more boring, more predictable, more safe, more wishy-washy, more vague. Is corporate banality the future of AI?.
23
58
274
@galo2099 It's a meta-analysis, so there's more than one study in the dancing category. I think you can criticize the result by pointing out that confidence rating for them is "low" but it's still a massive effect and there are other "low" categories without that.
7
0
254
How much of Claude's agency when playing Pokemon is like bowling with bumpers? After all, it turns out the literal number π can also play Pokemon.
11
7
271
2. People don't understand science is still (happily) a medieval institution. It's entirely about apprenticeship. To become a scientist, your first goal is to become a useful apprentice. Here's my guide (and tour of how science actually works):
9
28
262
3. The ultimate cause is that they are the result of what happens when culture is generated unconsciously. An unconscious culture is necessarily an exercise in high strangeness.
3
14
261
Going over the letters of Princess Elisabeth of Bohemia—mathematician, philosopher, eventual abbess, and the possible romantic interest of Descartes—who may have been the first person to fully understand the paradoxical nature of the mind-problem.
8
44
244
My magnum opus of the fall: the arguments for scientific incompleteness and consciousness being a true but unprovable statement within science.
27
41
257
I'm not the first to point this out, but here are some very cool stories of memories across heart transplants that I found as part of this.
Non-neuronal cells look increasingly like neurons, with the same memory pathways. Meaning that nutritionists are confused for the same reason neuroscientists usually are: your body is a mathematical black-box, with its output your morphology.
9
24
261
This blew past my promise of publication on 50 RTs, so here it is today:. Einstein > von Neumann because a sense of beauty and intuition is more important to advancing science than a high IQ (yet our academic "meritocracy" misses these traits completely!)
18
38
251
THIS! YES! Everyone from Elon Musk to Yuval Harari to Yoshua Bengio are calling for a pause on making systems more powerful than GPT-4 so that we can establish they are safe. Huge names on the list. I added my signature. Everyone on Earth should sign!
102
41
238
IIT is already being changed to be tagged as "pseudoscience" on Google as well as "pseudoscience" on Wikipedia as well.
Yesterday a letter from 100 scientists declared the popular theory of consciousness - Integrated Information Theory - is "pseudoscience.". And that media about it is "scientific misinformation." . The letter is bad. Here's my reply why:
16
16
230
On why Sam Altman really got fired: it wasn't about effective altruism, nor secret breakthroughs. It was that the board felt they caught him in a lie.
21
16
226
In the last week alone, the effective altruist movement has been on the cover of The New Yorker, The NYT, and Time Magazine. It has billions in funding, and wants to make the world a better place. The problem is that it's poisonous. 🧵
25
31
218
Why I expect writing, thinking, and art to get worse following Gen AI, not better
37
33
218
Scientists aren't usually open about this stuff, so I will be: a lot of this came to me in dreams. I dreamed some systems were like the rocks carved by wind, ventifacts, top-heavy with macroscale structure.
🚨 New paper, and big personal news! 🧵. First, I just published a new theory of emergence. It traverses the dimension reductions of a system, treating each scale like a 2D slice of a 3D object, looking for what each adds causally and irreducibly.
13
19
229
I feel like every famous physicist has metaphysical nigh-spirital views about the universe (e.g., Roger Penrose) and then the more normal but less august physics community has to work really hard to not acknowledge their beliefs
19
11
214
@loobah_l Someone else questioned exactly that too - how much is driven by contentious divorces.
2
0
217