Will it end with a paperclip maximizer? Or a permafrost refrigerator

you know the thought experiment by Svedish philosopher Bostrom, teach an AGI to maximize paperclip production and it will fagocitate the world unless you teach him/her/it ethics, human “values” and not values

But then you discover that an oil compay want to refrigerate the ground in order to save the permafrost that thaws due to fossil emissions https://grist.org/energy/the-irony-conocophillips-plans-to-freeze-thawing-permafr

imagine how much permafrost in the norther hemisphere is foundation to roads, houses, plants, pipeline, lines. Here is a map from the good The Arctic by Dodds and Woodward

so here you have a perfect doomsday machine without having to resort to mithical AGI’s, perfect economic logic of nowadays oilmen and, executives and politicians. Pump more oil in order to keep pumping oil, pump some more to keep houses standing and roads open, fossil emissions will raise temperatures even further, ramp up oil production and keep going to 5 degrees warming, and pumop more to 10 degrees warming and then checkout what’s left of our civilization

The Permafrost Refrigerator is a thought experiment by a 1500 pound goat, not a swedish philospher, though it sounds more real than the paperclip maximizer


It taskes 1,000 neurons (tech) to model a neuron (rat)

Deepmind built a DL model with 5 to 8 levels and up to a 1,000 nodes in order to forecast with 99% confidence the firing of a single rat neuron

How Computationally Complex Is a Single Neuron?

“If each biological neuron is like a five-layer artificial neural network, then perhaps an image classification network with 50 layers is equivalent to 10 real neurons in a biological network”

“the authors investigated which features of the pyramidal neuron inspired the much greater complexity in their five-to-eight-layer deep neural networks. They concluded that it came from the dendritic trees, and from a specific receptor that receives chemical messengers on the surface of the dendrites

Cells with learning


“To Gelber, the experiments demonstrated that Paramecium was learning to associate the wire with food, a conclusion that challenged scientists’ belief that only highly evolved, multicellular animals with central nervous systems were capable of such behavio. More fundamentally, her results suggested that at least some of the biological machinery needed for learning and other cognitive processes might exist not in the connections among neurons in an animal brain, but within individual cells themselvesr”

Previously on this site “Neurons with memory”https://1500poundgoat.com/2021/04/21/neurons-with-memory/”

I see a pattern here, at least MR Tyler Cowen getting interested in alternative explanations to neurons connecting, I found both links on his website

Neurons with memory

here says an experiment wtih ferrets show that memories are probably stored directly in the neuron, interview with Randy Gallistel who stresses that it would be a paradigm change for neuroscience and also #ai both assuming that memories are the effect of connection between neurons

“The most interesting idea” the idea would be revolutionary, the experiment, if recognised and reproduced, could be an anomaly that neuroscience has to face and accomodate in current theory https://join.substack.com/p/is-this-the-most-interesting-idea

GPT-3 links

and some renewed interest in Transformers

Gwern https://www.gwern.net/newsletter/2020/05#gpt-3

HN comments where iI found the link https://news.ycombinator.com/item?id=23623845

some intuition on relation between Graph and Transformer architecture https://graphdeeplearning.github.io/post/transformers-are-gnns/

I should find a good intro to transformer, it seems that they scale the right way, even bigger models with ever better performances, like size drives performance.

UPDATE 24 may 2021: Are we in an AI overhang?

Viagra for growth

il Sildenafil è il Viagra, il Taldafil è il Cialis, il Vardenafil è il #meToo col brand meno forte, non ricordo quale

Ash Jogalekar@curiouswavefn chiede “Show me a molecular modeling or machine learning technique that predicts tadalafil (Cialis) from sildenafil (Viagra) without the former being explicitly represented in the training set, purely based on target information. That’s the kind of scaffold hop that will turn heads. “

la biochimica non è tutta così prevedibile, però la performance di Alphafold nel torneo di piegatura delle proteine ha lasciato una grande impressione

Già da un pò si pensa che iniettando un po di tech (leggi software e ML) nella ricerca pharma si può girare la #eroomLaw in #MooreLaw se non ci credete passate da A16z che ha cominciato a usare il termine Bio come buzzword analoga a Tech, tech non è più una abbrevizione di tecnologia, come Bio vuole essere una etichetta breve e riconoscibile di una tesi di investimento nel bioscience https://a16z.com/bio/

Solo che l’eccitazione del successo di Alphafold adesso fa pensare a qualcuno che anche la Great Stagnation possa essere giunta al termine, insomma Alphafold è viagra per gli economisti

Se lo chiede Tyler Cowen che il termine l’ha coniato nel suolibro del 2011. E un po di tech optimism si vede qua e là

2 letture per il giorno di festa, distopia e golpe

Noah Smith propone una distopia, plausibile ma secondo lui non la più probabile. Al centro sempre la Cina con il mix di autoritarianismo e tecnologia, il partito che si arma di intelligenza artificiale per controllare ogni piccolo evento della vita dei cinesi, allenando tutto sulla oppressione degli Uiguri https://noahpinion.substack.com/p/the-super-scary-theory-of-the-21st

Zeynep Tufekci sociologa USA di origine turca, lamenta che laddove in Turchia ci sono diversi termini per descrivere ogni tipo di golpe, in USA ne hanno uno solo e non riescono a cogliere in pieno il rischio del ridicolo comportamento di Trump dopo le elezioni, che giunge dopo anni di occupazione e imbastardimento della politica americana, gerrymandering etc.

Anche Napoleone III era un pagliaccio che fu portato al potere proprio per essere una macchietta, salvo poi prenderselo e farsi imperatore , per lui Marx disse “Hegel says somewhere that that great historic facts and personages recur twice. He forgot to add: “Once as tragedy, and again as farce.” https://www.theatlantic.com/ideas/archive/2020/12/trumps-farcical-inept-and-deadly-serious-coup-attempt/617309/

of Biochemistry and Geometry

some Nobel prize was awarded 50 years ago to someone who claimed a protein’s shape could be derived from the atoms building it, so it started a rece at guessing how a protein would fold, folding proteins in short. It came to artificial intelligence news this week, Alpha Fold of Deepmin won some protein-folding olympics https://www.nature.com/articles/d41586-020-03348-4

Twitter reminds me how important and difficult to deal with, is atom’s position in a molecule, detrmining properties https://twitter.com/curiouswavefn/status/1334562488495423489

while reading PIHKAL di Shulgin you can find a chapter named the 4-posiyion where it shows how halucinogenic potency derives from molecules taking the 4- position in a benzene circlae and stay there while in our body

Orthogonality, USA elections and AI

in my usa bubble on twitter, mostly founders and founder phylosophers, Robin Hanson is makeng rounds with his pos that advises To Oppose Polarization, Tug Sideways

Americans a re under shock after 2 boomers, worse, 2 silent generationers 1942 and 1946 babled at each other wthou much sense for 1 hour. Politics is a multidimensional tug-of-war and in certain cases it is better mind your own fight and tug sideways

orthogonality is a condition thay ensures that AIs need not obliterate us humans https://arbital.com/p/orthogonality/ found on the twitter of

fonder of the rationalisic community lesswrong and this nice arbital site with lots of tutorials

the purpose of adovcting otrhogonality is different

to Hanson when politics gets fsctituois better go orthogonal

AIs in Moravec dialogue with … think of him as too factitious and cannot bother more with the real 4-dim world, thay have gone their own multidimensional way

here we come at orthogonality is Y meaning, AIs neednot get confrontational with us, there’s plenty of orthonogal room 🙂