Mindscape Ask Me Anything, Sean Carroll | July 2023
Вставка
- Опубліковано 18 тра 2024
- Patreon: / seanmcarroll
Blog post with audio player, show notes, and transcript: www.preposterousuniverse.com/...
Welcome to the July 2023 Ask Me Anything episode of Mindscape! These monthly excursions are funded by Patreon supporters (who are also the ones asking the questions). We take questions asked by Patreons, whittle them down to a more manageable number -- based primarily on whether I have anything interesting to say about them, not whether the questions themselves are good -- and sometimes group them together if they are about a similar topic.
We're experimenting with a new benefit for Patreon supporters: short video (or audio-only) reflections by me on the podcast that just happened. If you've been wondering whether to join up, this could be the time.
Mindscape Podcast playlist: • Mindscape Podcast
Sean Carroll channel: / seancarroll
#podcast #ideas #science #philosophy #culture - Наука та технологія
24:37 "Sometimes its right to be rude" Proof that Sean is on his villain arc
Loved every second of it. than you Sean
Any chance we could get timestamps on the questions?
Three hours of heart loving enjoyment. Longing for August already. 😊 Thank you Sean. 👍
Sean Carroll is the goat😂😢😮
Can you get Sabine on?
Sean pls get David Deutsch on!
Chris langan?
I'd prefer david grusch
i watched this while sleeping, thanks
If you think I'm going to burn forever, you can't also love me. That's insane to me.
I allways enjoy this .
The discussion (at 20:13) of the Berkeleyan question of whether information exists absent perception reminded me of the perennial argument over whether mathematical theorems are invented or discovered. Anyway, Shannon's initial exposition of information theory was firmly seated in a context inseparable from the intentions of humans building and operating communications technology (summarizing a paper I read over 50 years ago). Organisms are voracious generators and consumers of information (DNA!), even when their perceptual systems are quite basal. It strikes me that when I learned in 8th grade that irritability is a property of organisms, I didn't know that 60 years later, I would restate that as "organisms respond to environmental information."
I do still have trouble with the idea of information as an a priori property of physical systems, although the characterization of it as opposite to entropy is intriguing. Oh, good: another topic to study 🤔
Reflection as a tier is an excellent idea
Sean Carroll might benefit from the understanding of mind and consciousness outlined by Joscha Bach neuroscientist. Love Mindscape -very very inspiring. Thank you
My fear of AI is not existential but rather economical as the corporations creating LLMs are not concerned about the jobs being displaced by them.
This isn't exactly true.
Source: an AI developer hired to train people afraid of being replaced with robots into people that watch and handle the robots
If there's one thing you can trust a large company to do is look for the cheapest option and uhhh training people is more cheap than getting rid of functional jobs from people that perform a complex function lol
If decisions need to be made, "AI" is not ready to handle this task. We really shouldn't be calling this stuff "AI" anyways... it's more like fancy averaging than a true intelligence
Tend to agree especially given all new jobs and industries AI will create will be prone to the same problem.. Even if the resulting unemployment rate is 20% (much more is estimated) it’s enough to destabilize our democracies..
@@aidenmurphy9924 agreed. It's actually fancy filtering. Not AI. We are nowhere near achieving AI. Maybe in 100 years, but I doubt then either.
@@triaxon3791I think that you're conflating "conscious awareness" with "intelligence". For although it's true that you won't find too many conscious agents that do not also display intelligence, it doesn't then follow that intelligence is similarly predicated upon the presence of a conscious agent. AI is already intelligent. And how long until an ANI reaches the point of an AGI? Well, it won't be 100 years. Try somewhere between 1 and 5.
But then mass unemployment has been looming on the economic horizon for some time now, and quite regardless of recent advances in AI. Most of the advanced Western nations have been actively engaged in attempting to sort out some form of Universal Basic Income (UBI) for over a decade now. I'm also agnostic as to the true state of existential risk posed by AI - it may or may not yet spell the end for humanity. According to the most recent polling taken from AI and machine learning researchers, the percentage of industry professionals predicting our imminent doom (30%) are still outweighed by the more optimistic among them, who predict we will somehow survive (70%).
That said, though, I doubt you'd find any such level of disagreement regarding the future of employment from among economists, who are all pretty much of the same view here...
The long working day of _homo sapiens sapiens_ is almost over, and that holds true whether we all wake up next Monday or not.
This is why I always open my pasta sauce bottles in a sealed glove box and send a sample to a lab for analysis. Yes, supper is generally quite late on pasta night, but you can’t be too careful.
That's why I use pizza sauce on my spaghetti.
@@JustOneAsbestoWise move.
@@cullenjohnson0 Life hacks.
I'm glad it's not just me.
1:46:55 I used the funnel with filter pour-over method of making coffee for a while. A maintenance worker had left some junk on the floor near the sink, and it necessitated me placing my feet in an unusual position. Because of that, I accidentally knocked over the pour-over apparatus, and the boiling hot water soaked my sock. I got it off quickly but not before my upper foot was burned. I ended up doing 9 days in the hospital having a skin graft.
Thus, I do not recommend the pour-over method of making coffee.
As always: interesting!
Sean for president.
Josh wants to know how far he traveled: If you multiply ONLY the 185 m/s of the cluster speed for one year, then the number you originally added up will be the amount of travel you made between the two galactic points. (you zig-zagged a lot between those two points/we all have traveled, so your original larger number is a long way taken)
Joe
Indianapolis don't
Square yes 1x3x4 yes 4 is -1 is 3 looking for how many lanes not?
So 4 is 1x3x4 is 12 is 12 lanes---
x1 is 1x3x4 or 12 is 1x11x12.☝️
I feel like the answer to the question about energy conservation was explained somewhat poorly and possibly even missed the point entirely...
The question as i understand it was "how can space expand if energy can not be created, wouldn't the expansion of space necessitate the creation of energy?"
And the simple answer is yes. This sounds like a problem if you were only taught about Conservation in grade school (which is fine, no offense intended) but it isn't actually a problem for a really simple reason: the law of conservation of energy is local. Meaning it's generally true in a given isolated space but is not necessarily universally true everywhere all the time.
There is a physical quantity called the energy momentum tensor (a tensor is essentially a set of equations that describes physical properties of things) that describes the density and variation over time of energy and momentum(generally represented as P). It is the basis for calculating gravitational field properties in GR. We say energy is locally conserved because in GR the calculation of the covariant divergence P (the change of the energy density and flux of a given volume of space time) is equal to 0. This means that when everything is accounted for the total energy entering and leaving a system is the same and thus the total energy of the system remains constant. So energy is conserved locally.
What that means is when you measure an arbitrarily defined system the only thing you need to account for is the energy flowing in and out of that system, not the total energy of the universe as a whole. The reason we don't say energy conservation is global is that it is essentially impossible to define the total energy of the entire universe and there is no reason to assume that it would be constant. This is because it's impossible to define the total volume of the universe and as a result any attempt to map the energy momentum tensor to any arbitrary region of spacetime produces a different answer. To deal with this you would need some way of measuring and comparing the total energy and flux of different regions at different points in time, a universally true energy symmetry.
TLDR: we can't say conservation is global because we can not even begin to assess the total energy density and flux of the universe and even if we could there is no definable physical reason it would have to remain constant. This is a because it isn't possible to know the exact volume of the universe due to the fact that parts of it become incapable of interacting with other parts over time (they get too far apart too quickly).
I think the statement "there is more than one definition of energy" is true but kind of esoteric. It's easier to explain that the definition of what energy you're measuring can change. Imagine if i asked you to measure the total volume of water in a bath tub but the bath tub is actively draining and someone is filling it with buckets of water. Any particular total volume measurements would have to be a measurement of a specific moment in time and would become invalid immediately. So measuring total volume over time is essentially impossible in any meaningful sense.
So how do we know the amount of water being bucketed into the tub is equal to the amount draining out at any given time? We don't. We can't. Thus we can not say conservation is global.
A Million for Patreon is what I feel like, but my bank is not feeling it.
I give a dollar na episode, 4$/month is nothing for 6+hours of content. It's the best value entertainment out there. Yes I could get it for free and for years I even did but now that I pay I feel even more invested.
That thumbnail...it kills me everytime.
0:00: 🎙 In this episode, the host introduces the ask me anything addition to the Mindscape podcast and discusses the potential risks of AI, the different types of logic, and the conservation of energy in the universe.
13:56: 🔬 The speaker discusses various topics including the misconception about the formula for electrical current, the possibility of dark matter being composed of low mass bosonic particles, the concept of information in physics, and the consideration of others' reactions when expressing thoughts.
27:00: 😕 The anthropic principle is more useful when considering a multiverse or different conditions rather than a uniform universe.
40:53: 🔬 Sean Carroll discusses the assumptions and implications of the Boltzmann brain thought experiment and the relationship between quantum computing and the many worlds interpretation.
54:18: 🔬 Scientists should engage in countering bad ideas publicly, but in the right way, to reach out to the middle ground of people who haven't yet made up their minds.
1:08:42: 📚 Sean Carroll discusses various topics including black holes, the beginning of the universe, moral philosophy, and the role of games in understanding advanced physics concepts.
1:20:42: 🧠 Sean Carroll discusses topics including scientific revolutions, jazz clubs, black holes, personal identity, and idealism.
1:33:29: :coffee: Sean Carroll enjoys good coffee but is not a coffee snob.
1:52:22: 🤔 The concept of self in the many worlds interpretation of quantum mechanics requires a more nuanced understanding, as there are multiple versions of oneself in different branches of the wave function.
1:59:04: 🕰 The question of whether information is conserved in the universe depends on the interpretation of quantum mechanics and the type of information being considered.
2:12:18: 😕 The conversation covers various topics including the concept of self, black holes, choosing a kitten, science funding, and Einstein's general relativity.
2:24:21: 🔬 The energy tensor at a point can cause gravity, and in the early Universe, the energy density in photons was higher than in matter particles. In an eternal inflation scenario, the boundary of the bubble of post-inflationary space propagates outward through inflationary space. The concept of a first cause is misasked in the context of the fundamental nature of reality. Teaching modern physics in middle school can be beneficial, but it should be done in a simplified and accurate manner. The acceleration does not slow down as you approach the speed of light, but time dilation occurs. The formation of groups like QAnon, flat earthers, and Maga is a complex phenomenon that can be attributed to the allure of simple and clear explanations, psychological desires, and the need for an in-group.
2:38:26: 🔬 In quantum mechanics, the measurement of a quantum system can change its energy, but this does not necessarily apply to measurements of angular momentum.
3:02:05: 👶 Using a Quantum Universe splitter to decide on the last name of their first child.
3:03:14: 📚 The speaker discusses the concept of multiple worlds in quantum mechanics and how it could affect happiness.
Recap by Tammy AI
(2:53:10) I have served on a jury and I think that there is a definite similarity to the pursuit of scientific truth.
As a jury we were presented with a number of different versions of the truth from the various witnesses. Each witness had his or her own set of biases. I would compare it with all the various types of telescopes that astronomers have at their disposal for observing the universe. Each one has its strengths and limitations.
The jury had to piece together the most plausible story given the various reasons to believe or disbelieve the various witnesses. In the end, as a juror one knows that there is only one truth and that you have to identify which of the witnesses are most likely to be giving you misleading information. Just as the astronomer has to decide which of his telescopes are likely to be providing inconsistent data. In both cases there’s only one truth. In both cases there’s always a lack of agreement.
This a very accessible means to a better understanding of how the world works (at least w/ Sean’s bias, which doesn’t concern me as my understanding matches his which seems very clearly based on the Scientic Method, ie, we observe, predict, test, observe, then confirm re refute). I consider an observation as a type of experience. I assume that we all admit that we don’t necessarily have sophisticated or powerful enough tools to understand many things that seem so or seem that they should be so, but that we will continue to invent more sophisticated or powerful tools to clear up more and more of the unknown and to fit into our understand understanding.
My difference may arise in that all our knowledge is filtered and defined by our own senses. We are never able to directly experience the world because the data feed to us is filtered by the limitations of our senses. It may be that though we create tools to allow us to sense more things than our native senses allow. Not sure how he manages that concern. Further I am open to the idea that we may never be able to create tools that get to the levels of sophistication and power that we need to explain all things.
The primary idea that is impacted by the prior paragraph is how we explain consciousness and understanding how consciousness arises. There is a significant amount of seemingly reliable data to indicate that consciousness may not arise from the brain and its complexity, ie, as the number of connections increases at some point consciousness then arises from the brain. There are some extreme cases of people with extremely limited brain matter yet function and behave just as those with complete brains. Of course there are also the many studies by respected researchers into near death experiences that have clearly correlated no brain activity with the time that no brain activity was observable in the cerebral cortex, thus refuting the hypoxic hallucination idea.
One can easily argue both sides of this, but so far there are no good proofs that show that consciousness arises from the brain. Nor are there completely acceptable proofs that the opposite is true. This, of course, remains one of, if not the, most difficult challenge for completing our understand of the world and our place in it.
In the meantime I will continue to learn and enjoy listening to these podcasts. Sean, thank you so much for your work at work and in creating these videos.
Sincerely,
InscrutableJohn
The item in that video is not acting like a superconductor.. it's acting like a permanent magnet. It flips over to jump to the favorable orientation and slaps down on the bottom magnet. Superconductors never do that. Permanent magnets do
Wrong video, ignore
Yes, gravity is "pushing you down", but it's an almost negligible effect. The wings on a mosquito can easily defeat the entire gravity of Earth. Rather, the reason that we "feel" as though we are being pushed down against the surface of the planet is that our legs are constantly preventing the surface of the planet from coming up and smacking us in the face! Well, our legs are preventing our face from going nowhere, and thus then being hit by the surface of the planet as it expands outwards into space! This effect explains why gravitational mass and inertial mass cancel each other out, and so why any two objects will always appear to "fall" at the same rate - it's because objects in free-fall aren't actually "falling"! When you let go of something with your hands it _doesn't move anywhere at all._ It stays _exactly_ where you left it in space, and it's the surface of the Earth that rushes up to meet it. Your feet feel tired at the end of the day because they've been supporting your body against the surface of the planet as it constantly expands outwardly.
But, if you still don't believe me you can try this experiment. Go up on your rooftop with a pair of scales strapped to your feet and see what you weigh. Ok. Got it? Good, now jump off the roof and have another look at those scales, you now weigh... nothing! Right? Bodies in free fall are inertial observers - they do not experience gravity because they are no longer accelerating. This is Einstein's "equivalence principle".
But hey, if you _still_ don't believe me then go have a look at Veritisiums video on gravity, or ScienceClics excellent video's explaining general relativity...
Or maybe just go take a short jump off a tall building. But, once you do _get_ it, you'll never see things "falling" in quite the same way again... it's a trip!
2:02:36 Well, if "everything is compatible with the eternalist view of time" ( i.e. the Block Universe) then the whole idea is rather insignificant , no?
If it is compatible with everything, then it says nothing in particular.
I am only at the AI - pasta sauce analogy(so I didn't hear any further explanation), but one counter to this analogy is that multiple countries and private corporations are not currently engaged in an arms race to intentionally weaponize jars of pasta, unlike AI.
I suggest Naomi Wolf as guest.
There is no intrinsic jiggliness to quantum fields....I do not know what it means but I like the sound of it....
12:00 - Professor Carroll discusses the stuff of our Universe. ^.^
AI: IF it's successful in the stock market? A single user, who notices the AI is increasing gains that are constantly positive ... will not share that info with anybody... they will let it go, and they win everything. EVERYTHING!! Just like the game Monopoly, that scenario would take all the money. Don't fall for AI-chosen stocks and portfolios. Remember what/how to think about chance, and "what would Sean Carroll" say on the subject. (big & small numbers/probability)
Everything in the above stated plays out to No money left for the 8 billion other people on the planet.
8:25 That's my opinion, too...
When we say that humans are intelligent,is this a theory?
To the Patreon questioner, if you subscribe to UA-cam Premium. You will get rid of all Google inserted ads across all UA-cam channels. There can still be "commercials" within some channels if the content provider chooses to film and insert them themselves. UA-cam Premium would not get you the content reserved for Patreon supporters.
Great, now I can't conceive of opening a can of pasta sauce.
I the time is relative so Kant had thoght right. Hadn't it!? More important than the TIMA is the space.
My TV has your face burned into it.
😂
I feel like not worrying about the existential risk at all would be silly of us as a species akin to not worrying about quantum mechanics ect.
11:20 its not correct to say that the universe must conserve every
👍👍👍
I heard an opinion about AI, where the danger is simply that a bad person can speak to a hyper intelligence and get ideas.😮
That kinda made me think.
Oh Sean, you need to try a stovetop percolator.
It's truly amazing.😊
I love my 30 year old PYREX glass percolator!
I can't figure out if your two comments are related. I imagine Sean speaking to a hyper-intelligent AI and gets convinced to try a stovetop percolator... XD
I think you will have scant luck convincing anyone who enjoys pour over, that a percolator provides any real advantage.
What is the probability of the existential risk of AI compared to that of consuming the contents of a jar of pasta sauce? Are they the same? I do not think they are?
In 5th grade my teacher said I had unique logic. It wasn't a compliment...
(Ref 9:50)
i can not afford to ask sean a question. however.. someone can and might like MY suggestions. a: "jacque fresco" and the "venus project" b: the content in lyrics of the band "bad religion"
Neutron decay cosmology is inevitable. Geometry is a harsh Mistress.
I am frustrated at the AI danger discussion going on in this episode. It feels like the discussions of climate change 50 years ago.
Only Einstein could see spacetime , even though it was deducible from the orbits of planets.
As we follow Moore's Law, the size of transistors are getting so small and numerous that there is a concern about quantum tunneling. My question(s) would be;
Would a series/parallel (+/-) design help mitigate this concern, or,
Could we predict at what distance and scale that tunneling might occur, and then use that in some way to our favor?
Parallel design would help, but it then runs up against another constraint, namely heat. A chip design using 3 dimensions would be subject to the cube-square law with respect to shedding waste heat.
I believe we already know the scale and are pretty close to it. Using it to our favor would require a complete new style of computing, probably quantum computing.
@@CorwynGCThanks for the helpful reply. I was working through how we've used perceived impediments to our advantage in the past by reframing our engineering approach or our application of energy to the system.
the part i don't understand is if the AI's have values .. wouldn't they be human values or am i missing something .. good podcast by the way
I believe that David Deutsch asserts this, though unfortunately I can't find where I saw that.
OMG! You think it will be that bad? If you create a new human then their brain is created like everyone else's, but some of them become greedy corporate heads, tyrannical dictators, or serial killers. Since we are approximating how to create a human-like brain then they will probably be at least as varied. They may be far more moral and view the average human as fake and shallow.
They may be human values, but not necessarily a complete or balanced set of human values. Like other synthetic entities - corporations, for instance - they could value some things too much, and others not enough. From Bhopal to climate change to the Titan, if it's in a corporation's or AI's short-term interest to kill us, it will not hesitate.
AI doesn't exist. Everything out there is simply a machine-learning model. The companies call it AI because it sells, but it's NOT AI. Humanity is so,,, so.. so far away from AI that we haven't even defined "intelligence" for ourselves, let alone transferred that to a machine.. Worry about human values another hundred years from now, maybe. (probably not even then)
But what are human values exactly? Can you distill them into some mathematical function
Is a seed higher or lower entropy than a tree? Couldn't the universe have been sown?
If someone did sow the universe , then who sowed them ?.
i've seen both sides of the spectrum, happily dated a very religious person and went to church with them and everything and i also have seen a bible thumping militant atheist(accidentally of my own creation..stupid video about praying to a milk jug...) which can be worse than a preacher! if you dont believe exactly what they believe, they go into a tirade and dont listen to anything you have to say and become as closed minded as a monk who believes if he doesnt open his eyes every morning, the world doesnt exist for anyone but him and he tries to delay the world existing by walking around with his eyes closed most mornings
Teacher get paid quite a bit more than your average profession. If we wanted better teachers we'd probably fire 90 percent of them and replace them with industry professionals. At the very least, teaching in and of itself should not be a college major. It should be a minor you take with a field of study. If you want to teach math you should be a mathamatician with a teaching minor. Give our teachers that level of expertise and I'm fine with making them more. But not as it stands now, no.
AI: I've asked AI to write me the #1 song/novel/movie... and so far it's failed me miserably. It has also obviously failed everybody else the same. It is a tool, and use it as such. If you're older, then find somebody young who uses it well. Tie your horse to them. Embrace wisdom tied to new inventions.
Isn't math a language? Why not train AI trained to do and understand the patterns of math and physics and chemistry?
Seems if AI were truly smarter it would have solved math and physics problems. Like how to build a fusion reactor.
Tbh AI doesn't seem actually intelligent to me.
It's just really good at rephrasing old Info.
I was good at that in 2nd grade using my encyclopedia to write essays to look smart in school.
I wasn't. I just am good at plagiarizing.
Spoiler alert. This comment relates to subject matter late in the podcast. . . . Compared to mathematics and quantum theory being depicted in the movies, Sex and violence are both way over represented. I guess the screen writers REALLY don't want the general audience practicing advanced mathematics or particle physics. I guess the powers that be have decided we are safer with a public that is punching each other and having orgasms than thinking deeply about anything.
2:28 I disagree. I think the AI threat is greater than we can acknowledge.
I know Sean warns that many of us are anthropomorphizing intelligence, but actually it is Sean anthropomorphizing intelligence.
We have AI over a barrel but only for the next 6 to 12 months.
If you can shoot me with a water hose, but I can kink the hose to stop you,
then it doesn't matter that you are twice as smart as me.
But if you are 200,000 times smarter than me, then my ability to kink the hose doesn't help.
Why not ?
I don't know. I"m only 1/200,000th smart enough to answer that.
I defer to Eliezer Yudkowsky's TED talk : ua-cam.com/video/Yd0yQ9yxSYY/v-deo.html
that came out yesterday. Then see his Lex Fridman talk.
He's been studying AI and AI moral alignment for 22 years.
He has nothing to sell and no one to impress.
He is much more informed on this.
It is the only both pessimistic and fatalistic TED talk I've ever heard.
His TED talk is only 5:45 since it was hurriedly requested by TED.
I watched that talk. He IS selling what he believes: he invented and worked on it for over 20 years. He's talking about "super intelligent AI", which is a complete fantasy. He's worried about something we don't have/doesn't exist.
AI doesn't exist.
A pet cat is more "intelligent" than ChatGPT. Just because an app can put together words that seem like a human wrote them isn't intelligence. It is a complex digital filter, just like a Google search, or search on your PC, or a computer chess game.
Don't fall for the salesmanship. AI is a marketing phrase.
Eliezer believes in the god he discovered 20 years ago, but that god has not come down to earth.
BTW: the title of the TED talk uses "super intelligent artificial intelligence" in its title, (not very intelligent title)... and do you notice the audience is laughing and smiling?
TED only gave him 5 minutes, and the ovation at the end was for the host walk on.
@@BrianFedirko Anything that hasn't happened yet can't happen.
I'll try to see things in that light from now on.
All these institutions, corporations, universities and enterprise start-ups who are all fighting tooth-and-nail and investing billions to develop the most powerful AI and AGI are all just mistaken.
Why ?
Because you say so.
With AI in the news this much, you must be cramping up from sighing, shrugging and rolling your eyes.
We need more experts like you.
@@BrianFedirko Wow. You've explained everything. When the audience laughs, it must be ridicule.
They couldn't be laughing due to the out-of-place sense awkwardness ?
Eliezer comes in with his unremittingly pessimistic ted talk at TED itself (not just a TEDx talk) which is known for its relentlessly optimistic can-do spirit.
Couldn't be that.
And yes, I'm sure they're giving a stand-up ovation for the host who is always just hanging around there and comes out periodically.
Y'know, right at the moment Eliezer finishes his talk and thanks the audience.
Could be.
@@BrianFedirko "X hasn't happened yet, we need not worry" - wow, a truly genius statement.
Not related but Jesus is coming back please get right with God before it’s too late
2000 years and still no evidence. Read the room before saying silly things.
Oh god, what the hell did Rogan do now? There was a new thing?
Edit: Okay, yeah, I looked it up. The whole Bud Light ad thing. Good ole Joe "I don't hate trans people but they just pick their gender every morning when they wake up and then spend all day sexually abusing children" Rogan. Yeah, great job not being hateful, buddy.
Where did he say anything like this
Sean is providing three (3) hours of content that will need to last you exactly thirty (30) days. Your happiness is unwaveringly dependent on whether you are actively consuming fresh content from Sean. When you are NOT consuming fresh content from Sean you are unhappy. If you are unhappy for seven days (7) consecutively you will fall into an irreversible docuseries binge.
What is your ideal daily routine in order to maximize your happiness for 30 days?