AI rediscovers Einstein's Time Dilation and Kepler's 3rd Law

[@TylerJosephson-x6u]

Co-author of the AI-Descartes paper here - thanks for featuring our work on your show!

[@m.t-thoughts8919]

You are one hell of a cool mad man. Alot of respect for that work 💯

[@TylerJosephson-x6u]

@@azurnxo2134 There are many possible paths I’ve seen others take, and all involve learning math (typically up to calculus or more) and programming. I started as a chemical engineer (where we learn both math and programming, plus chemistry and more), and I started learning AI because AlphaGo was really exciting. I attended a symposium on machine learning at my university, and I met someone from IBM who eventually introduced to his team, which led to our collaboration and this paper. So in addition to learning fundamentals in math and programming, getting excited about it and learning more on your own, you should also practice networking and try to connect with those working in the field.

[@vascojoao]

@@TylerJosephson-x6u Marketing, how to 'sell' your paper, that's what you have to learn, @AzurnXO . I'm a dropout of Computer Engineering, and I can tell you, after Einstein everything is wrong with Theoretical Physics, just like Sabine has pointed too. Maths, it's another game, not a rigged one. Only Nash tried to play 'tricks' on Maths and we all know by now that 'self-interest' it's the stupid form to progress on a closed System. AI is basically a lot of Flip-Flop's together feeding 'data' to the previous entry, Intelligence for sure it isn't maybe that's why they've called it Artificial.

[@ralboraggins9564]

@@TylerJosephson-x6u me me me me me me me me me me. everyone look at me

[@patford9943]

Congratulations on 6 months of “Science News”. Your presentation and explanations are spot on. Keep up your good work!

[@DarkForcesStudio]

Have only discovered this channel recently. So glad I did. You've reignited my love of science. Physics has lost it's way to some extent in recent years. This honest and no nonsense approach respects the intelligence of its audience. Thank you Sabine.

[@HauntedHarmonics]

Same! She’s an excellent, to-the-point science educator. No filler or popsci BS. So glad I stumbled on her channel recently

[@DarkForcesStudio]

@@HauntedHarmonics Yep. Also the dry humour seems to go over a lot of heads. A great find.

[@maalikserebryakov]

I believe in Dark Matter

[@DJWESG1]

There seemed to be a moment when everyone was sad and depressed because they all thought the universe would just eventually die out in a cold vast nothingness.. But I think that depressive perspective has passed. The notion of multiverse theory has taken hold and people are beginning to see the self reproducing potentials in the theories and observations..

[@eljcd]

Welcome, I thin you could enjoy her blog bacreaction. Sadly she doesn't publish anymore, but it has covered the last 15 years in Physics with the trademark style Sabine

[@ReynaSingh]

The engagement with these videos is outstanding. Keep it up Sabine

[@UniDocs_Mahapushpa_Cyavana]

I agree.

[@smlanka4u]

The medium changes the speed of light. And slowing down causes the speed dilation called time dilation. Therefore, the time dilation concept is only a concept.

[@smlanka4u]

@@_John_P, Prof. A. Einstien's principles can't make time relative. The flow of time continues from moment to moment in everywhere, making the relativeness.

[@Andytlp]

@@smlanka4u einsteins theory of relativity works. Its proven. If u want to reinvent the wheel youll have to try harder. As the guy above suggested. Gpt explains it pretty well and the experiments: The bending of light: One of the key predictions of relativity is that gravity can bend light. This was tested in 1919 during a solar eclipse when astronomers observed the positions of stars near the sun. The sun's gravity bent the light from the stars, causing them to appear in slightly different positions than they would have otherwise. The results matched Einstein's predictions. Time dilation: According to relativity, time can appear to move slower for objects that are moving very quickly. This effect has been confirmed in many experiments, including one in which atomic clocks were flown around the world on jets. The clocks on the jets ended up ticking slightly slower than those on the ground, just as relativity predicted. GPS: The Global Positioning System (GPS) relies on relativity to work properly. GPS satellites are moving very quickly in space, and their clocks are affected by time dilation. If this effect were not taken into account, GPS would be off by several kilometers.

[@itskarl79]

Just curious why Sabine ignores the 100ppm-5000ppm CO2 range that we can live in, and the fact that we are currently 400ish ppm, which is a lot closer to the bottom end. Meaning if the CO2 is reduced by a couple hundred ppm, all plants begin dying (means we all starve shortly after)... Seems like everyone bought off on that dipshit Bill Nye because he was on Saturday morning kids shows... I agree with the damage caused by refrigerants, but we haven't pushed to change the handling and recycling of any products/machines that use it... Why do climate people misrepresent plant food?? (CO2) It's also coming from a constant rate of decay of the Earths crust with crust movement/volcanic activity, and a bunch of other stuff known for 60 years. When are we going to let the science do the talking and not the corporate funding that bully's scientists into pushing false or exaggerated parts?

[@ponyote]

Congratulations on 6 months of this. Very glad to have such a great source of news from a presenter I trust.

[@PlanetEarth3141]

Did Sabine say six months?

[@inventorbrothers7053]

These science news videos are gold!

[@rand49er]

There is a level of thoroughness in the topics presented that I can't get anywhere else, and it's right here on my laptop! Thank you so much, Sabine!

[@Luxamor-8]

No, no allowances made here on innovation with the metals needed for energy transition. Sodium Ion is already a thing, with Sodium 50,000x as prevalent and NIFe magnetic replacing rare earths.

[@TheGhostPariah]

Only 6 months?? I definitely loved this format from the beginning! Thanks for doing what you do :)

[@Manorainjan]

If You have already difficulty to remember that it was only 6 months, how likely is it, that You remember the complex content that was presented?

[@GEMSofGOD_com]

What's the freaking thing about this telephone though, I don't get it

[@laughy38247357075834]

​@@GEMSofGOD_com same. I hate the telephone gag

[@Manorainjan]

@@dcm621 "It's a colloquial way of complimenting the channel." Really? When You are at the dentist, and he tells You that to fill the hole will take only 5 minutes, then "only 5 minutes?" is a compliment.

[@Manorainjan]

@@GEMSofGOD_com "the freaking thing about this telephone" has at least 3 purposes: 1) Sabine can bitch about any very stupid political matter in an indirect way. 2) She can associate herself with the extraordinarily famous Elon Musk in the hope that his glory will rub off on her. 3) Like all kinds of running gags, it adds entertainment, loosening up and strengthens the identification of the onlooker with the program.

[@ut4321]

Love, love, love this program! Weekly science news from a top physicist who also has a sense of humor? Yes please!

[@Ikus13]

This is the best science news outlet out there today. Your objectivity is very refreshing and trustworthy. Congrats on the 6 months!

[@Naptosis]

And that Sabine is very open about any mistakes and corrects them. Instead of putting corrections on page 19 in 8 point font like the legacy media does. Sabine's a great educator!

[@Houshalter]

There was an AI from the 70s (!) that rediscovered Kepler's formulas given only the data he would have had access to. This evolved into something called symbolic regression. And software for it has been available for a long time

[@jimmythebold589]

but did it have a GUI?????

[@Superman-Tube]

That's true, the big difference is that the AI-Descartes also derives relations from background theory (e.g. conservation laws, invariants, symmetries) using logical reasoning, very much in the way that a theorist would approach the problem. If the relations are not derivable from the background theory (say the background theory is lacking some relations, or some of which are incorrect), the algorithm would quantitatively assess the distance of the symbolic hypotheses from any implicit derivable relation.

[@ThePowerLover]

@@Superman-Tube This.

[@localverse]

​@@Superman-Tube so we have to feed the AI all other equations and theories except for one, and it has to figure out the missing one? (I'm trying to get a sense of how the AI works)

[@Superman-Tube]

@@localverse The framework in the AI-Descartes publication, at the extremes can ingest anything between no background theory, and complete set. For the former it will purely rely upon symbolic regression (which isn't new, yet may provide hypotheses in symbolic form, which fits the data to a certain extent). In the latter case, assuming that one of the hypotheses that was brought forward by the symbolic regression module is fully derivable, the system will provide a certificate of the hypothesis derivability. In the more likely case where we are somewhere in between, that is we have some partial background theory, AI-Descartes will assess (quantitatively, rather than in a binary form) the distance of each hypothesis from the background theory. So one can qualify models not only by data fitness and complexity, but also by how derivable they are.

[@P-G-77]

Quantum Theory - General Relativity... i added many data, equations and i ask for hypotheses based on the interpretation of IA data, IA logic, formulating various hypotheses about how quants interact with matter and always hypothetically, using mathematics to describe these hypotheses… the results were… strange to the start, absurd at a certain point, then the equations appeared... I saved everything for future use. Asking precise, logical questions that can always form "hypotheses" on topics that in any case must be described exactly, even with equations, continue the discussion and move forward step by step, always by hypothesis... and often these hypotheses become very "possible"

[@chippysteve4524]

Great work Sabine! Love the new format. You are an exceptional science communicator.

[@todhannigan8779]

I look forward to weekly sci news! Best shortform newscast period. You also list your sources, which is great for continued researching! Thanks!

[@peramoredellanalisi4341]

From science to popular scientific news. Good job, Sabine.

[@Taricus]

I was part of a research team at one point, synthesizing carbon nanotubes to do that very thing. I was also the guy that used the scanning electron microscope to check the samples of carbon nanotubes, so we could improve the process. ^~^ The reason it works is that you have nanoparticles in the nanotubes, so they will rotate in the magnetic field. Then, the rotation causes the cancer cell to go through a heat death, but leaves normal cells alone, because the cancer cells are targeted directly.

[@russellsnyder2634]

I am so glad you're doing this! Back in the 1970s, I used to read Science Digest. But then they changed after a few years and made most of their articles about carbon emissions and climate. I quit them when they published a long article about "what if dinosaurs had survived?".You cover everything.

[@Llortnerof]

Goes to show just how long we've known about that particular issue and how long politicians have been dragging their heels, though.

[@johnscaramis2515]

Regarding electric vehicles: many of those metals would also be needed in combustion engines. Like e.g. cobalt to desulfur fuel. Or other metals for catalytic converters. And the amount of metals is reducing constantly: cobalt is not a necessity in batteries (LiFePO4 batteries) and CATL will bring a natrium/sodium based battery to market (also no cobalt) in the near future. Other metals are necessary not for the batteries, but for the electric motors (sometimes people throw both things together, sometimes deliberately). However you could use asynchronous motors instead of synchronous motors to avoid the use of high-strength magnets.

[@levmatta]

Go brasilians scientists!! Congrats. It is really hard to do science in Brasil. Parabéns, orgulho de vocês.

[@dougbriggs6797]

@4:08 - I haven't read the paper and this might be a minor quibble, but I was taught that when resistance increases then the current drops to minimum while the voltage goes to maximum. Is it different in plasma physics ? Congratulations on your 6 month anniversary !

[@davidriosg]

oh, Sabine, your videos are so good and your sense of humor and jokes are impeccable. Thank you for the great content!

[@brianorso3504]

That opening song is very retro, I like it.

[@jessicathompson2491]

An efficient delivery of useful and fascinating information, thank you for this!

[@emtechproaudio6176]

I love it when you work your sense of humor in. The funny bits are smart and done with a dry style that really tickles! More!

[@michaelblacktree]

Wow, it's been 6 months already? But I'm glad you decided to do the science news. It's one of my favorite things on youtube.

[@gefginn3699]

Great post Sabine. I appreciate all the information you are sharing here. I always enjoy tuning into your newest post.

[@kalrandom7387]

Shutting off nuclear will be the biggest mistake we can make

[@jake_]

"We've put the soup cans together with the cereal boxes to grow one, too." Golden.

[@parallaxnick637]

“And to this end they built themselves a stupendous super-computer which was so amazingly intelligent that even before its data banks had even been connected up it had started from base principles with "I think therefore I am", and got as far as deducing the existence of rice pudding and income tax before anyone managed to turn it off.”

[@susanlippy1009]

Yes the answer to life is 42🤣

[@dougbamford]

Thanks for 6 months of news. There are now sodium ion battery chemistries that use more common materials. It seems likely that smaller cars will end up with sodium and only heavy/expensive/performance cars will end up with the premium batteries. Small cars and better public transport are the answer, not massive cars with huge expensive batteries.

[@Blabla130]

1:13 "They developed an algorythem that works like a theoretical physicist to work. I mean someone's got to do it" In theory.

[@belledetector]

Sabine, your Science News is a great concept, entertaining and educational. I truly hope you will keep it up.

[@cybervigilante]

This week: AI discovers old science. Next week: AI discovers new science.

[@NobleUnclean]

Yes!!

[@andreasf8170]

The news about the rare metals remind me, that in the late 19th century the Times in London predicted that by 1950 the streets would be covered with a ten foot layer of manure. 😮

[@ParniaSh]

I love this series! I've learned so much from you.

[@rigorobles3991]

I was compelled to come here and write my comment :) excellent video, would love to see more.

[@wihlke]

I'm glad you and Elon get along so well, Sabine. So happy I introduced you two!

[@bjornolson6527]

Very well; snidely so? 😊😮😂

[@simateix6262]

Science news have quickly become something I'm looking forward every week

[@MariaMaria-nl3xc]

este vídeo tem legendas em português e espanhol, Obrigado Sabine!

[@rararnanan7244]

Whenever I click on Sabine’s science news the 1st thing I ask myself is “will the phone ring?”…

[@moresoysauce5489]

Thank you Sabine for reigniting my love for science. I am a daily viewer and I am almost finished with your book. I also subscribed to brilliant haha. Thank you so much for doing what you do ❤

[@bannor99]

April 17th was International Bat Day so good one, Sabine

[@agnosticmuslim6341]

You are the best!!! Thanks for being entertaining all the while informing at the same time!

[@jeffneptune2922]

Thank you Sabine. I really like these concise episodes that allow people interested current scientific research to keep up with some of the noteworthy studies.

[@kapsi]

I wonder what the Descartes AI says about measurements of galaxy rotations, galaxy clusters and expanding universe

[@CAThompson]

Dark Matter versus MOND?

[@kindlin]

Great question! I have a feeling, tho, that there are more biases in that model than we think (in all models) so it would probably just spit back out what we already think, like it's already done, but we can't know until we try.

[@LastMotivateUs]

That is brilliant work, keep up like this

[@robertsoyka1822]

About the Tesla cybertruck: "No, no it's a great car. We've put the tin cans together with the cereal boxes to grow one, too." 🤣Made my day...

[@notanemoprog]

MDS

[@MiqelDotCom]

and he's so over-sensitive to criticism that he called her immediately, lol.

[@mickeyfilmer5551]

René Dustcarts... " I Stink, Therefore I am!" (British sense of humour! 🙄🤣)

[@nealwright5630]

Love your videos, Sabine! I'm a bit dubious about the Net Zero goals and dates. I have a feeling it is going to be next to impossible to get full compliance on these goals. There will be a lot of resistence.

[@russbell6418]

Especially from totalitarian governments who see their own progress as the only progress.

[@felicityggreene7831]

The goals were unrealistic in the first place. Studies like ones she mentioned should have been done before the goals were set, but they weren't. The only practical way to ensure we have the raw materials to do it are to exploit poor countries more than we already are. Do you support a military intervention to secure mining rights?

[@my-tschischlak]

very intresting, especially the cancer part! Thank you every week for crawling science news and sum up in a video 👍

[@DJWESG1]

Good news for sure.

[@edreusser4741]

I keep wondering what the widespread adoption of iron nitride batteries will do to all these numbers. I hope that when they start replacing lithium-ion power sources, the manufacturers are smart enough to produce form-matching batteries. Iron nitride doesn't take any rare earth metals or other expensive materials to make work. The maximum field possible is almost 4 times larger than the maximum amount possible from N52 (the current best boron-neodymium magnet material). The current manufacturer is realizing a 30% higher field than the N52 maximum, so it has plenty of room to grow.

[@scribblescrabble3185]

same for sodium and lithium.

[@LewdCustomer]

The battery metals are recycled with 90% recovery. So mining will increase for the EV transition, then decrease when enough dead batteries can be turned into battery metals.

[@dreadwinter]

Germany shutting down all nuclear power plants is a completely ignorance-based over reaction, right?

[@Thomas-gk42]

Have to have a look in th history, to understand

[@Nebulus42]

Atoms are bad . No poison for our children

[@skeltek7487]

It would had been more impressive if Keplers law was deducted by just giving it the angles of stars as viewed from a telescope.

[@shadowdragon3521]

Machine learning being able to derive physical laws from observational data is very exciting because it means that we could soon see a monumental leap forward in our scientific understanding. I can't wait for the day when AI comes up with a completely new theory all on its own.

[@toomanycharacter]

It'll come up with its own equations, and that won't mean much until we understand how and why those equations are true. The AI in question won't generate a comprehensive theory of everything in a way that humans understand, it will instead generate equations that fit previous observations well.

[@GeneralELL]

@Cirit 42

[@Superman-Tube]

​@@toomanycharacter that's a very important point. The novelty of the AI-Descartes work is exactly addressing your concern. Rather than merely fitting the data with spurious relations, the algorithm, ingests all background theory known (say laws of conservation, invariants, symmetries, other axioms), and qualifies how far a given hypothesis is from any derivable form. When the hypothesis is fully derivable from theory, you will get the derivation path (as we all appreciate from physics classes and books), and when the theory is incomplete (say one omitted conservation of angular momentum,.or newton 2ns law from the background theory), the algorithm will assess the distance of the hypotheses from any derivable form. The attempt here was to bring formal logic to the automated discovery world

[@toomanycharacter]

@@Superman-Tube I stand corrected.

[@hansolafsen77]

I'm afraid it won't be possible to write the law down in a set of closed formulae. Maybe it makes good predictions but it won't explain anything. But maybe there is an AI capable of translating it to scientists' language...

[@MrHichammohsen1]

Great to see you explain sun physics with elctromagnetic terms. Hannes Alfvenn would have been so proud.

[@JohnnieHougaardNielsen]

Did the AI deriving a formula out of planetary motion also look closer at Mercury?

[@jsl151850b]

0:52 *As seen in the movie 'Colossus: The Forbin Project' where the US computer figures out a new theory of gravity after it inexplicably speeds up.*

[@Smidge204]

10:58 Gonna see if I can find that report later and see if they made any allowance for modern EV battery chemistries that don't use cobalt, or the use of motors that don't use rare earth magnets (platinum group metals). These are materials that are really beneficial but not strictly necessary, and auto makers (at least outside of China) are shying away from them exactly because of the political issues. There's also plenty of lithium, it's just a matter of how much we're wiling to pay to get it. There's literally billions of tons of Lithium in the oceans but getting it is... challenging.

[@davidh9638]

Why is lithium such a problem as lithium batteries are replaced with sodium batteries?

[@Smidge204]

@@davidh9638 Well for starters nobody manufactures Na-Ion batteries yet so it's not a great idea to pin your hopes on a tech that functionally doesn't exist yet. Stories about new battery technology are cheap and plentiful but until you can actually buy one it doesn't count for sh*t... For two Na-Ion batteries have lower energy and power densities compared to Li-Ion so they are not a 1:1 replacement. It's not clear how much Na-Ion can displace Li-Ion in practice until the tech matures and the actual performance goes from theoretical/in-a-lab to something real.

[@russbell6418]

At 6 months, I’m already hearing people say (about various scientific press articles) “I’ll wait for Sabine’s opinion on it”.

[@mr5384]

Hi Sabine! Love your videos. There's a news article going on about photos being generated from gravitational waves. Would be interested in your coverage!

[@michaelginever732]

Making predictions is difficult; especially about the future. Not sure who said that, but they were right. Predicting the amount of various resources that will be needed to electrify transport is subject to that difficulty. Already the need for lithium would seem to have dropped - as has the price of the lightest of alkali metals. The reason being that CATL has gone into production of a Na ion battery that has a respectable energy density. Na of course is far more abundant than Li and so way cheaper too. As for cobalt, which indeed is mined in the Congo by child labour in appalling condition, Li Fe PO4 batteries solve that. They don't quite have the energy density of Li, Ni, Co, Mn batteries but they are safer, will maintain capacity through more charge / discharge cycles and tolerate wider ambient temperature range. As for magnets in motors there have been some developments there too recently which reduce the need for Nd. There are solutions, there are talented people working on them. We need to get the corrupting influence of fossil fuel money out of politics. With the will we will be able to transition. We have started several decades too late but there is now real hope. And yes Germany has made a mistake turning off all their nuclear reactors before they have the renewables to replace them. They've done it now so let's hope they get into high gear bringing on the renewables.

[@Lathamhipsurgery]

Love this channel and the way that you present such interesting topics. I’ve seen patients whose symptoms improve temporarily after an MRI scan. It’s probably due to tissue heating. Using magnetic fields to activate intracellular nanotechnology has huge potential in the treatment of tumours. One day maybe we’ll be able to ‘electrocute’ cancer by harnessing this energy!

[@jonthompson6673]

Thank you for the stock picture. I didn’t know what someone with brain cancer looked like.

[@clipmixhd4937]

Sabine, can you please make a video about efuels the same way you did on hydrogen? I would really like to see your opinion on them considering it has been in the German news a lot lately

[@SabineHossenfelder]

It's on the list already!

[@odizzido]

There is a decent video made already, but it has a bit of a narrow focus and we could certainly use more videos on the subject kzbin.info/www/bejne/haGoc2meeNV8fa8

[@clipmixhd4937]

@@SabineHossenfelder Thanks so much! (Hätte ich auch auf deutsch schreiben können, aber ich möchte nicht so heraustechen :D)

[@eyeofthasky]

@@clipmixhd4937 u did already by stating its in the german news a lot -- only germans follow german news for the most part :D

[@VoodooMcVee]

@@clipmixhd4937 Harald Lesch made a video on that topic some days ago. But I'm still curious to hear what Frau Hossenfelder has to say about it.

[@chadbailey3623]

I would love it if you would include links to the studies you share with us.

[@rael5469]

There's the phone. I can't wait for it to ring.

[@robertpearson8798]

AI rediscovering scientific laws is all well and good until one of them comes up with the answer 42 or begins to suspect that there are some extremely odd things about the mission.

[@sparky7915]

Excellent video as always! For me it is mind boggling how people can study objects at such great distances. Incredible! Nanotubes are an exciting and great way to treat brain cancers. Maybe they treat other ailments with these nanotubes.

[@Naptosis]

What jab? Nature leaves virus particles in our bodies permanently all the time. I'm sure the tubes will have a method of removal or a way to make them benign.

[@mikespangler98]

If you have an incurable cancer the risk from the nanotubes is tolerable.

[@Llortnerof]

@/O/ Or maybe *you* should look up the controversy behind Campbell instead of advertising somebody known for spreading misinformation about the topic and just blabbering complete nonsense about homeostasis that only shows you have no idea what you're even talking about.

[@Llortnerof]

@/O/ And i have no tolerance for those that abuse statistical data to try and promote conspiratorial viewpoints. So shut up and get lost.

[@DaveJablonski]

I really liked the intro graphic and music.

[@Kyedo2022]

Problem is that nuclear power incidents are like airplane crashes, suuper rare but devastating when it does happen. That being said I would build a house next to a cooling tower and get my spa water from it also. :)

[@Debrafeem]

Yet nearly 10 million people die every year from air pollution induced by the burning of fossil fuels. Not devastating at all…

[@scribblescrabble3185]

That is actually an argument _for_ nuclear power. A catastrophic failure means a huge area of land won't be settled or used by humans for a long time, which means it is guaranteed to be an S-grade biotop for many generations to come. Turns out human interference is more damaging to life than radiation.

[@ultraranger1286]

Yes, devastating not just in terms of physical damages to properties, environment and lives but also to public opinions. For example, airplane is statistically the safest way to travel, much safer than cars who kill much more people per year. Yet you don't see 3 hour long documentaries and non stop media coverage of car accidents. Same with nuclear energy. Fossil fuels kill much more people and do much more damage to the environment per year through air pollution, mining accidents, drill leaks etc. yet every time you mentioned nuclear they can't stop whining about Chernobyl and Fukushima.

[@danilooliveira6580]

except modern power plants are basically impossible to meltdown though. and even if we consider older power plants, they technically only went wrong once. three miles island was basically nothing, mostly mediatic panic, and Fukushima even with the entire deck stacked against the plant, it needed the biggest tsunami in history to make meltdown, what could have been easily avoided if they only listened to the warnings of professionals.

[@gregmattson2238]

@Steven F - they are only 'devastating' because we make them so. One person died - potentially - from the fukushima power disaster, wheras an estimated 2000 people died because they were moved from otherwise safe areas. The real culprit is the ability of homo sapiens to assess risk.

[@GiacobbeWillka]

This could be groundbreaking for unsolved equations with the use of AIs and its potential of automation. Scientists and animators could really make use of generative AIs like Bluewillow to name a few.

[@blackshard641]

I've been saying for years that it's entirely possible the next "Albert Einstein" or "Steven Hawking" will be an AI. That's not only seeming increasingly likely, but at the rate things are advancing I'm starting to think the breakthroughs might start coming within the next year or two.

[@brandongonsalves3615]

I went to a lecture at Stanford 2 weeks ago and the department of energy is building GPU’s there and at Berkeley for AI to find “antibiotics for everything”.

[@Llortnerof]

They might help humans find solutions to known problems, but i rather doubt they're going to come up with anything new. It'll be a long time before they can do anything without human interference.

[@MAXpower10000000]

Thank you for your great weekly recaps🙌 love to hear from you !

[@_DATA_EXPUNGED_]

F_cking Magnets, how do they work?

[@dragons_red]

Batteries

[@kx4532]

Yo! kzbin.info/www/bejne/f5qao5p_hsaqodE

[@korenn9381]

Ask the AI! I'm sure it'll be able to work it out soon.

[@ooglyga6100]

that an insane clown proposition.

[@markbarton8872]

Hello Sabine, my first comment. The report that you presented today April 18th regarding precious metals in battery technology was using old data. CATL, the largest battery manufacturer in the world based in China, is already producing sodium based batteries with a tiny amount of precious metals in them. Chery based in China is already producing cars with CATL sodium batteries. Cost, politics, and general availability have forced the largest battery companies to find alternative materials that are both plentiful and inexpensive. CAN'T WAIT FOR YOUR EPISODE ON ELECTRICAL INFRASTRUCTURE.

[@randalllaplante358]

I assume that searching for unknown physics would be an upcoming task for AI. I wonder how it would do it, and when will it/they start (if they haven't already. Bonus pun: Is Anti-De Sit-ter space the same as Standing space? (Which would be useful in a theater.)

[@UniDocs_Mahapushpa_Cyavana]

If they were given time inputs, they might be able to detect *changes in the* laws 📜 of physics.

[@davidrandell2224]

“The Final Theory: Rethinking Our Scientific Legacy “, Mark McCutcheon for proper physics.

[@orbatos]

Not unknown physics, consolidated models of existing observations and theories. Remember, none of this is actually ai.

[@tizqar]

Edison -> Wayne -> Musk -> Tesla (Einstein) Looking forward to see more :)

[@rolandrickphotography]

That Neutron Star thing at 8:01 is quite interesting. Isn’t it something like a very exotic gigantic nucleus of an atom and for that also a gigantic quantum mechanical subject? Just doing some brainstorming… 😊

[@brianmcdaniels8249]

Jon Evans, eidetic/hyperthemasia guy, yea him, discovered quite a lot, including the non expansion of the universe, no darkmatter, and the whole cosmological problem top to bottom, plus a ton of other things.

[@dazley8021]

Germany is insane for turning off nuclear power in a time of crisis like this. And switching to more coal power ontop of that? Wow, i'm immensely disappointed in this country and their awful decision making.

[@Thomas-gk42]

The idea is, to fill it with wind and solar, should be given a chance

[@tgdomnemo5052]

We had enough nuclear accidents, high time to switch off. ... the coal thing though 😮 It's politics, sadly.

[@bjornolson6527]

Nuclear-based energy was a gift, and still is. There are newer, scalable, and safer options in this space. Let’s just pledge not to dump the spent waste in the ocean. Rather, continue to research better/safer methods of disposal. We can do this!!

[@Stefano_SPGamer]

totally agree

[@pensive8552]

Good job AI! Now do a unified field theory!

[@aqdrobert]

Terminator: Vee shall destroy all of you with a unified field! JA!

[@irgendwieanders2121]

Maybe if we called it a unified front theory, more people would get behind research?

[@pensive8552]

@@irgendwieanders2121 🤦 😑 get out 👉 lol ♥️

[@hansolafsen77]

Already in the 80s, Herbert Simon's AI BACON, did that already. It builds increasingly complex mathematical relations, and tested them and came up with Kepler's third law. And this approach was "interpretable" as well. It's strategy to come up with the "Formula" was understandable by humans as well.

[@yrunaked4]

I love it when she says Einstein 😀

[@fred_2021]

Einshtein? ikr :)

[@daveking-sandbox9263]

She’s German, of course she knows how to pronounce Einstein. She laughs at the way you say it! 🥳

[@stephaneclerc667]

That juice stuff was gold

[@trolloftime5340]

Hopefully physicists don’t become obsolete because then i have no career paths

[@dragons_red]

They already are. The field has been stalled for decades. Look at becoming a prof, engineer or go into an engineer related field as a backup. I graduated with Physics degree but went into an engineering related tech field. You will make way more money too.

[@trolloftime5340]

@@dragons_red I’m in engineering rn, will go for computer science 2nd year if possible, but I only realised I wanted to do physics after like 6 years of prepping to go into cs through school. Doing cs mostly because it will give me the foundations for physics later down the line and a good wage if things don’t work out.

[@yingyang1008]

You became obsolete when you started creating concepts like space time

[@nicholashylton6857]

Someone's gotta do the experiments. Reality doesn't have to conform to any human's or AI's expectations.

[@trolloftime5340]

@@yingyang1008 bro talks while having a profile picture related to balance between light and darkness and all that

[@ratking1330]

A counter point to the electrical vehicle paper, it's under the assumption that there is no innovation in battery technology and continue to use lithium based batteries. Sodium batteries for example are made up of very common elements, so manufacturing cost and material sourcing will be much cheaper in conparison to lithium ion. Depending on how fast nano-tech progresses as well, we may see the rollout of graphine batteries as well. Personally I would lean toward graphine. Not only is it more energy dense than both lithium and sodium based batteries, but since it's purely electrical rather than converting chemical energy to electrical, there isn't risk of combustion in an accident for example. Though I believe you've done a video on batteries. (Edited, had incorrect info)

[@jccklh]

Sodium batteries are less energy dense, but are cheaper, so a good solution for grid storage, and reduce the need for Li.

[@ratking1330]

@John Craig I thought sodium batteries were more energy dense. Did quick looking into it and from what I gather, they hold similar amounts of charge (lithium does indeed have the edge) but sodium batteries last 3x longer and work under a greater range of temperature. Though sodium batteries appear to be heavier than lithium if you go off of atomic mass alone. In practice though, I am unaware of the weight difference

[@sophiophile]

​@@jccklh There are companies that have produced sodium ion batteries with equal energy density at lab scale to the lithium ions used commercially.

[@zhelmd]

Physists are getting obsolete 😮😮😮

[@bartroberts1514]

It's okay that the phone rings. It's very Get Smart. But maybe something more updated? Could the smartphone ring?

[@Thomas-gk42]

She's a bit nostalgic

[@starbird14]

Here is a modified equation for the third law as follows: P^2 = (4π^2/G) * (a^3 / M) where P is the period of revolution, a is the semi-major axis, G is the gravitational constant, and M is the total mass of the central star and all other objects in the system.

[@trikk9964]

... you are great, Sabine!

[@juvenalsdad4175]

Excellent as usual. On the subjects of nuclear power and transport, I have still not seen a cogent argument as to why using SMR's to power synthetic hydrocarbon fuel production is not a better option than electric vehicles.

[@IanSlothieRolfe]

Its almost like moving away from Nuclear power is a move in the wrong direction... How could that be?

[@WilfEsme]

This is one of the good thing about AI, if it's included in the dataset, it can perform most scientific algorithms. Except those that we believe are misinformation, it's unable to prove this type if it's wrong. I think any physics related computations could really be done through AI. I hope that one day, image generators like Bluewillow AI could prove to be useful in other sciences.

[@barstopcu3207]

I just need to write, sabine the Work you do is amazing. Thank you and keep going

[@brownro214]

Great report Sabine. Keep 'em coming.

[@sidefx3]

This reminds me of a show that was on when I was young called Beyond 2000 or Beyond Tomorrow.

[@BrunoEggert]

'Rare earth' metals means the group of lanthanides (the row of La-Lu). Li, Ni, Co, or the noble metals mentioned in the video are not rare-earths.

[@ofskittlez]

It's a treat to see you on Wednesdays as well as Saturdays!