"Complexity" is a hugely problematic term when used in this way - remember that entropy and complexity are related, but they are not interchangeable. A complex system can have lower entropy than a simpler system, and conversely, a system can have high entropy but be relatively simple. By mingling these terms without specifying objective reference points, it all just comes out as word salad.
This paper just reads like an attempt at sounding smart while actually saying little.
> a system can have high entropy but be relatively simple.
Good examples of these are anything that Kolmogorov-compresses well. For example, by almost any measure the output of a pseudo random number generator has high entropy. Yet it has low information density (low complexity), as the program that generates the sequence, plus its state, is really small.
I think a better example is just hot gas. Heat up a tube of gas, and its entropy will increase, with no effect on its complexity. Still not terribly compressible either though.
Yes indeed. As I understand it, entropy is about states that are more likely.
I wonder if it always increases though? Eventually there will be enough entropy that any change may cause it to reduce or oscillate? (At universe / reachable universe scale).
It always increases in an isolated system. That caveat is almost always missing in pop-sci level of discussions about entropy, but it is crucial.
> Eventually there will be enough entropy that any change may cause it to reduce or oscillate?
Assuming that the universe is actually an isolated system, entropy will reach a maximum (it cannot oscillate). It is interesting to speculate, and of course our theories are imperfect and we are certainly missing something. In particular, the relationship between time and entropy is not straightforward. Very roughly: is the entropy a function of time, which we could define otherwise, or is time a consequence of entropy changes?
In the first case, we can suppose that if the universe reaches an entropy maximum we’d be far enough outside the conditions under which our theories work that we’d just have entropy decrease with time (i.e., the rule that entropy increases with time is only valid close to our usual conditions).
But in the second case, it would mean that the universe reached the end of time. It could evolve in any conceivable way (in terms of the fundamental laws of Physics), and the arrow of time would always point to the same moment. "What comes after?" Would be a question just as meaningless as "what came before the Big Bang?"
In any case, there are a lot of assumptions and uncertainty. The story does not do the subject any justice.
Yes, we call that state "heat death". Note that the second law is actually that entropy never decreases; it's allowed to stay constant for certain interactions (for instance I'm pretty sure an elastic collision preserves entropy).
The generations in my family are full of people who had dementia/Alzheimer's/auto-immune disorders. They all had anxiety, and they all had sleep disorders - I'm certain there is a connection among it all.
I joined a Stanford study and had a spinal tap sample submitted for testing - in theory I am "unlikely" to develop anything for the next 20 years - but that only gets me to my late 60's :(
From what I've found, being overweight negatively affects sleep. As far as I can tell, the extra weight simply makes it hard to breathe at night, leading to apnea. This can then cause depression and anxiety, depression because you wake up tired and feeling awful every single day and anxiety because you may wake up with night terrors after suffocating in your sleep.
At least it's easy to get a home sleep apnea test and an APAP/CPAP these days.
There are discords where users offer each other advice. That said, many of the tuning parameters require the doctor to set them and if you make an appointment you can talk over the problems and get them to adjust the pressure or whatever.
> That said, many of the tuning parameters require the doctor to set them
No, they don't. At least not on a Resmed machine. You just have to enter a special mode to change them. On my machine that's holding down 2 buttons for a couple of seconds.
> if you make an appointment you can talk over the problems and get them to adjust the pressure or whatever.
I'm glad you had a Dr that did this for you. It is not the universal experience, and it definitely wasn't mine.
Configuring the machine with anything other than the default settings
A new machine with the default settings is a nightmare. The apap algorithm will spike your pressure up to 16 (which is too high for most people) one minute and drop down to 4 (which is too low for most) the next. At the high end is aerophagia and leaks, at the low end, your apnea events will still happen.
When I called my doctor and explained the machine made my sleep worse, her office said “we gave you a machine. It’s out of our hands. Would you like us to refer you to a psychologist? The problem must be in your head.” This is almost verbatim.
The equipment supplier wasn’t much better. They offered to do a “mask fitting” but it was mid 2021 and the woman didn’t even want to be in the same room as me for fear of covid.
So what should they have done? I dunno. Eventually a friend recommended a mask that worked for me. I learned to analyze the data reported by the machine and changed the settings myself to something that worked for me. It took a year, the worst year of my life. Tired and constantly waking up. Thank god for YouTube and OSCAR. knowing what I know now, I can (and have) help someone else do this process in a week or two, but I had nobody to even ask.
I think it would be a good case for handing off from the prescribing doctor to some kind of home health aid / respiratory therapist who can make house calls and help make adjustments at the home.
It needs extensive calibration, if you don't want it to either be ineffective, or wake up in the morning, with your stomach pumped full of air, like a balloon.
> At least it's easy to get a home sleep apnea test and an APAP/CPAP these days.
If one wants to rule out sleep apean (being not the best sleeper) but probably doesn't have it (but other causes like stress) but at least want to measure it somewhat, how could one test that oneself with at least some probability?
Lots of things that are called "at home test" seem to involve getting special equipment and sending stuff to a lab anyway, that still involves a physician, some visitations with them etc... anyway.
Is there anything at all that one can do on their own to have at least some idea of the chances of having sleep apnea or not that does not involve a professional? Even if to then consider a professional if there are some indications? E.g. from an all-night audio recording, any indication of sounds that are yes or not sleep apnea indicating? And heartbeat/oxygen etc... measured by a watch?
Or is sleep apnea such a subtle and hard-to-detect issue that it's really impossible to self diagnose, it's such a subtle thing that only specialized equipment and specialists can do it, and trying to measure anything yourself gives no more answer than a 50/50 random chance coin toss?
The tester they give is basically just a pulse oximeter that logs your heart rate and SpO2 all night. I'm not sure how they interpret it, but I'd imagine that if your heart rate goes crazy and your oxygen drops, that's probably bad.
Otherwise you can look at questionnaires that ask things like whether you wake up a lot at night or snore, etc.
That said... it's pretty easy to just get a device that they give you that you connect to your phone to log data for a night. And they know what they're doing when they interpret it. So I would recommend that, since it's really easy to just go to a place and get a device and do a test.
This isn't like it used to be where you'd have to sleep in some strange place.
The at home tests require a quick before and after video call to a physician, and equipment mailed to you and mailed back... It's really not a big deal.
> anxiety because you may wake up with night terrors after suffocating in your sleep.
I have SA and a history of "mild" nightmares going back decades. Nothing terrifying, but very uncomfortable such as snakes hiding everywhere, and recently other animals such as tigers. I can't prove it, but I believe these are subconscious efforts by your brain to scare you into breathing.
I mean, panic attacks where you hyperventilate are basically the drowning reflex going off. So given that a lot of the dreams in our case were literally suffocating (drowning, etc.), I'm pretty sure that some part of the brain gets awakened enough to realize something, if not exactly what.
Same, except we also have digestive issues. I'm really hoping a paleo diet and good probiotics will save me. It's already helped me a lot. I'll let you know in 40 years if it prevents me from getting Alzheimer's.
[x] Highly stressed pretty much always due to work
I can only hope that whatever brand of dementia I end up with is something akin to Mr Toad's Wild Ride, or whatever. I'd be happy just to go out either laughing or well-rested.
If you’re supplementing probiotics, there is some research about inulin to talk through with your doctor. There’s a chance that inulin is carcinogenic even at lower doses. A lower dosage might be 4g and a typical probiotic is 200g.
Reasons to be wary of this research, including:
-The rodent-to-human study element
-causation v correlation
-sample size
Reasons to take this research seriously:
-Probiotic supplements are relatively new and long term effects aren’t understood.
-The “health as a business” realities that are very much at work here.
This is a strange comment, inulin is not a probiotic, it's a soluble dietary fiber. It's at best considered a prebiotic. Probiotics are microorganisms.
Culturelle contains inulin. That's the brand that takes up most of the space at my local CVS and Walgreens. Also the CVS-brand knockoff contains inulin.
...in miniscule amounts. Culturelle has 200mg inulin.
From the case study in the GP
>The only noted lifestyle change between colonoscopies was the initiation of 4 grams of agave inulin powder daily in his morning beverage for the last 2 years of that seven-year period.
I bet it would be outrageously expensive to replace. Escalators aren't cheap to begin with and this one is no doubt built to its own standard so anything you replace it with would have to be heavily customized to fit.
Alan Kay has promulgated many famous truths about computer science - one of them being that among all fields of study, it has the least regard for the people and discoveries that brought it to where it is today. Maybe it's just my own sense of history as I move into my 4th and likely last decade of working with computers, but I find this to be both true and lamentable.
How many of the people who made the steam engine possible do we remember? James Watt of course, but many many people were making contributions in material science, needed to make them useful. Not to mention many advances in values. No doubt lots of other areas as well, but I'm no expert in the steam engine.
Not sure what you mean by this - it's not as if steam engines are an extensive technology today, and certainly no university is teaching "steam science", while nearly every school is teaching "computer science".
Perhaps this is just the attitude that drives Mr. Kay's point home - do individuals who are interested in CS have little value for who and what has come before them?
To be fair, everybody who takes a thermodynamics course owes 90% of it to the 'steam science' pioneers. Understanding what determines and limits the efficiency of heat engines was as big a deal in its day as the WWW is in ours, but unlike our own era, a lot of brand-new science and math had to be discovered by those engineers.
Steam tech is much, much more interesting than it appears at first.
Sure they do, it's just called fluid dynamics and thermodynamics. Enough to fill up about 1/3 of a mechanical engineering degree, if you're really into it (most aren't).
Not that there's a lot of historical context to things as far as which people did what - most of that sort live on in names of techniques and methods (Rankine cycle, de Laval turbine, Carnot efficiency, etc.)
I - like ~20% of the humans of planet earth - have a bridged LAD coronary artery. The vast majority of those who have this congenital deformity will never know it and remain asymptomatic; a small cohort of those will, like me, have a minor heart attack and fully recover - a smaller cohort will die of sudden cardiac arrest, many of them males in their teens while practicing football or performing similar cardiac stressing events. Unfortunately determining this congenital deformity requires invasive radiological imaging - hopefully this technology will assist in it's identification sooner.
This is like claiming a photorestor controlled night light "understands when it is dark" or that a bimetallic strip thermostat "understands temperature". You can say those words, and it's syntactically correct but entirely incorrect semantically.
The post includes this caveat. Depending on your philosophical position about sentience you might say that LLMs can't possibly "understand" anything, and the post isn't trying to have that argument. But to the extent that an LLM can "understand" anything, you can study its understanding of nullability.
People don’t use “understand” for machines in science because people may or may not believe in the sentience of machines. That would be a weird catering to panpsychism.
Where is the boundary where this becomes semantically correct? It's easy for these kinds of discussions to go in circles, because nothing is well defined.
That depends entirely on whether you believe understanding requires consciousness.
I believe that the type of understanding demonstrated here doesn't. Consciousness only comes into play when we become aware that such understanding has taken place, not on the process itself.
Shameless plug of personal blog post, but relevant. Still not fully edited, so writing is a bit scattered, but crux is we now have the framework for talking about consciousness intelligently. It's not as mysterious as in the past, considering advances in non-equilibrium thermodynamics and the Free Energy Principle in particular.
Or like saying the photoreceptors in your retina understand when it's dark. Or like claiming the temperature sensitive ion channels in your peripheral nervous system understand how hot it is.
This is a fallacy I've seen enough on here that I think it needs a name. Maybe the fallacy of Theoretical Reducibility (doesn't really roll off the tongue)?
When challenged, everybody becomes an eliminative materialist even if it's inconsistent with their other views. It's very weird.
You declare this very plainly without evidence or argument, but this is an age-old controversial issue. It’s not self-evident to everyone, including philosophers.
It's not age-old nor is it controversial. LLMs aren't intelligent by any stretch of the imagination. Each word/token is chosen as that which is statistically most likely to follow the previous. There is no capability for understanding in the design of an LLM. It's not a matter of opinion; this just isn't how an LLM works.
Any comparison to the human brain is missing the point that an LLM only simulates one small part, and that's notably not the frontal lobe. That's required for intelligence, reasoning, self-awareness, etc.
So, no, it's not a question of philosophy. For an AI to enter that realm, it would need to be more than just an LLM with some bells and whistles; an LLM plus something else, perhaps, something fundamentally different which does not yet currently exist.
Many people don't think we have any good evidence that our brains aren't essentially the same thing: a stochastic statistical model that produces outputs based on inputs.
Of course, you're right. Neural networks mimic exactly that after all. I'm certain we'll see an ML model developed someday that fully mimics the human brain. But my point is an LLM isn't that; it's a language model only. I know it can seem intelligent sometimes, but it's important to understand what it's actually doing and not ascribe feelings to it that don't exist in reality.
Too many people these days are forgetting this key point and putting a dangerous amount of faith in ChatGPT etc. as a result. I've seen DOCTORS using ChatGPT for diagnosis. Ignorance is scary.
Do biologists and neuroscientists not have any good evidence or is that just computer scientists and engineers speaking outside of their field of expertise? There's always been this danger of taking computer and brain comparisons too literally.
If you're willing to torture the analogy you can find a way to describe literally anything as a system of outputs based on inputs. In the case of the brain to LLM comparison, people are inclined to do it because they're eager to anthropomorophize something that produces text they can interpret as a speaker, but it's totally incorrect to suggest that our brains are "essentially the same thing" as LLMs. The comparison is specious even on a surface level. It's like saying that birds and planes are "essentially the same thing" because flight was achieved by modeling planes after birds.
For example, they are dismal at math problems that aren't just slight variations of problems they've seen before.
Here's one by blackandredpenn where ChatGPT insisted the solution to problem that could be solved by high school / talented middle school students was correct, even after trying to convince it it was wrong. https://youtu.be/V0jhP7giYVY?si=sDE2a4w7WpNwp6zU&t=837
> For example, they are dismal at math problems that aren't just slight variations of problems they've seen before.
I know plenty of teachers who would describe their students the exact same way. The difference is mostly one of magnitude (of delta in competence), not quality.
Also, I think it's important to note that by "could be solved by high school / talented middle school students" you mean "specifically designed to challenge the top ~1% of them". Because if you say "LLMs only manage to beat 99% of middle schoolers at math", the claim seems a whole lot different.
> Each word/token is chosen as that which is statistically most likely to follow the previous.
The best way to predict the weather is to have a model which approximates the weather. The best way to predict the results of a physics simulation is to have a model which approximates the physical bodies in question. The best way to predict what word a human is going to write next is to have a model that approximates human thought.
LLMs don't approximate human thought, though. They approximate language. That's it.
Please, I'm begging you, go read some papers and watch some videos about machine learning and how LLMs actually work. It is not "thinking."
I fully realize neural networks can approximate human thought -- but we are not there yet, and when we do get there, it will be something that is not an LLM, because an LLM is not capable of that -- it's not designed to be.
> LLMs don't approximate human thought, though. ...Please, I'm begging you, go read some papers and watch some videos about machine learning and how LLMs actually work.
I know how LLMs work; so let me beg you in return, listen to me for a second.
You have a theoretical-only argument: LLMs do text prediction, and therefore it is not possible for them to actually think. And since it's not possible for them to actually think, you don't need to consider any other evidence.
I'm telling you, there's a flaw in your argument: In actuality, the best way to do text prediction is to think. An LLM that could actually think would be able to do text prediction better than an LLM that can't actually think; and the better an LLM is able to approximate human thought, the better its predictions will be. The fact that they're predicting text in no way proves that there's no thinking going on.
Now, that doesn't prove that LLMs actually are thinking; but it does mean that they might be thinking. And so you should think about how you would know if they're actually thinking or not.
Artificial neural networks already are approximating how neurons in a brain work, it's just at a scale that's several orders of magnitude smaller.
Our limiting factor for reaching brain-like intelligence via ANN is probably more of a hardware limitation. We would need over 100 TB to store the weights for the neurons, not to mention the ridiculous amount of compute to run it.
Language can be a (lossy) serialization of thought, yes. But language is not thought, nor inherently produced by thought. Most people agree that a process randomly producing grammatically correct sentences is not thinking.
That argument only really applies to base models. After that we train them to give correct and helpful answers, not just answers that are statistically probable in the training data.
But even if we ignore that subtlety, it's not obvious that training a model to predict the next token doesn't lead to a world model and an ability to apply it. If you gave a human 10 physics books and told them that in a month they have a test where they have to complete sentences from the book, which strategy do you think is more successful: trying to memorize the books word by word or trying to understand the content?
The argument that understanding is just an advanced form of compression far predates LLMs. LLMs clearly lack many of the facilities humans have. Their only concept of a physical world comes from text descriptions and stories. They have a very weird form of memory, no real agency (they only act when triggered) and our attempts at replicating an internal monologue are very crude. But understanding is one thing they may well have, and if the current generation of models doesn't have it the next generation might
The use of highly anthropomorphic language is always problematic- Does a photo resistor controlled nightlight have a chain of thought? Does it reason about its threshold value? Does it have an internal model of what is light, what is dark, and the role it plays in demarcation between the two?
Are the transistors executing the code within the confines even capable of intentionality? If so - where is it derived from?
As someone who lives with this reality moment to moment (thanks to a deeply bridged LAD coronary artery) it is interesting to see a visual representation of these effects.
It's incredible the lengths humanity is willing to go to avoid adopting nuclear energy - despite the US navy driving mobile reactors millions of miles over the last 70 years.
Reality is solar isn't viable everywhere. And it's not optimal to put it in places where you use the sun to grow food.
We should follow a holistic approach.
* Wind where it's windy.
* Solar where it's sunny - ideally on buildings/away from farms.
* Hydro where possible.
* Nuclear where it makes sense, i.e. stable geography, low occurence of natural disasters, lots of land.
* Some natty gas plants for overflow - not saying commission new ones or prioritize natty, but it's sensible to utilize existing peaked plants.
I'm not a big fan of large scale battery storage solutions, but they can work sometimes. I think they're more sensible for residential/commercial use and, when paired with solar, can really help add robustness to the grid. But, for mega energy storage, I think hydro based solutions are more sensible.
What do you base all these claims on? Plenty of papers show Solar + wind + storage is viable practically everywhere.
Also there is a ton of research on planning energy systems and what technology mixes make sense. This stuff has to be economical. Energy costs are measured in percentage of GDP. Simply liking nuclear doesn't make it viable. Especially in a world with PV meaning you can't sell energy during the day.
There really is only one macro fact that will shape the energy system of the future: The price of PV modules is now effectively zero in rich countries. Everything else has to be judged by how well it complements/makes use of free energy during daylight hours. The geopolitical implications of this haven't even begun to be explored.
Energy independence and a mix of technologies isn't just about cost. It's about redundancy and how you're positioning yourself to handle various unexpected events.
You don't want a state primarily on solar if you get a super cell darkening the sky for a week.
You want a mix of renewables, but you don't strictly want to rely on the food graces of mother nature at all times.
Boiling things down to just price is a very simplistic view.
You seem to think this contradicts my statement somehow?
Obviously you want a tech mix, and obviously anyone working in the field is taking the dunkelflaute or other extreme events very seriously. That's where storage comes in, and that's where the biggest unknowns and needs for future development are (e.g. is seasonal H2 storage really feasible). But to pretend like nuclear can magically become cheaper through technical breakthroughs, while storage is an unsolvable problem is disingenuous.
It's also disingenuous to suggest that random fluctuations in weather are somehow a unique problem. Sudden unscheduled maintenance can take down nuclear plants as well. As can the weather: Nuclear power plants require cooling and can be shut down due to weather and climatic conditions, too [1].
> I analyse climate-linked outages in nuclear power plants over the past three decades. My assessment shows that the average frequency of climate-induced disruptions has dramatically increased from 0.2 outage per reactor-year in the 1990s to 1.5 in the past decade. Based on the projections for adopted climate scenarios, the average annual energy loss of the global nuclear fleet is estimated to range between 0.8% and 1.4% in the mid-term (2046–2065) and 1.4% and 2.4% in the long term (2081–2100).
> But to pretend like nuclear can magically become cheaper through technical breakthroughs, while storage is an unsolvable problem is disingenuous.
I didn't say storage was not solvable and I even gave a better storage solution than your silly "batteries" example.
> Based on the projections for adopted climate scenarios, the average annual energy loss of the global nuclear fleet is estimated to range between 0.8% and 1.4% in the mid-term (2046–2065) and 1.4% and 2.4% in the long term (2081–2100).
From your own linked article - do you think this energy loss is even close to comparable to solar for similar conditions? You've linked an article but don't seem to understand the point they're looking to make.
Anywho, I don't think you're looking to argue in good faith and seem to have an anti-nuclear agenda, despite talking about an "energy mix". Save your policies for whatever echo chamber they were derived from, thanks.
If seasonal storage is solvable, solar + wind is not unreliable.
And no, I don't have an anti nuclear agenda. But I know the energy system models and the results, and just how difficult integrating nuclear into the mix is.
Finally, I know exactly what the paper says but maybe you don't: the problem with Dunkelflaute events is correlation. If it's cloudy somewhere and the sun is shining elsewhere, then no problem. These problematic conditions for nuclear are the same: large scale spatial correlated.
> Finally, I know exactly what the paper says but maybe you don't: the problem with Dunkelflaute events is correlation. If it's cloudy somewhere and the sun is shining elsewhere, then no problem. These problematic conditions for nuclear are the same: large scale spatial correlated.
Great. We'll just pipe over the energy from Arizona to Michigan, should be fine.
Large scale weather events drop nuclear by 1% long term. What percentage do they do for solar?
I'm not even a solar hater - I love solar... On residential and commercial rooftops. Or in sunny and void of life areas.
I love a mix. And nuclear integrates just fine into the mix. Look at a province like Ontario where 60% of the energy is derived from hydro and nuclear. An incredible and robust baseline power with low downtime and, correspondingly, cheap power for all of the residents of that province.
Even more energy from solar and wind too, with some natty gas as top off. Seems to work just fine for them - and with long winters and plenty of cloudy days, solar as a big component of their energy mix seems pretty silly to push for. You can see the mix live below.
It was 53% nuclear and 28% hydro at the time of me posting this. 16% natty, 4% wind, and 0.1% solar. The solar was good for 21 MW and the nuclear was good for 9600 MW for perspective. Their nuclear has been safe as hell and has run flawlessly for I think 30+ years.
PV is not heavily subsidized compared to other energy carriers, and the learning rate has been extremely consistent for a long time.
LCOE takes some of the system costs into account, but it's of course true.
And your second sentence is not how energy markets actually work. Of course I make a contract for reliable supply, but I don't contract with an individual power plant. I contract with an energy company and that buys from the cheapest supplier mix (aka merit order).
The long term contracts for base load run for years, but those, too will eventually have to adpat to the reality of abundant cheap daylight energy.
My main point with PV isn't about the system we have right now. It's that we are in the first days of a new system structured around the new technological reality that only recently emerged. Until very recently nobody, even the optimists, expected PV to get that cheap that fast. It will take decades for the repercussions of this phase transition to shake out. The issue is, due to climate change we don't have decades.
Great question! Mostly, because I forgot. Second, it's pretty expensive for the gains you get - I'd rather favour it for use in heat pumps; but it makes sense for some regions.
In what way is it not? It has a budget and a mission. Economics are a huge part of their existence. Far more is spent on health care every year over the Navy.
The US Navy has other, very high priorities to balance against cost - national security, global security, freedom, peace, trade, the lives of billions, the lives of its own personnel. The US military may the largest budget of any organization in the world.
At the same time, they invest in those mobile nuclear reactors only for specific needs: attack submarines, ballistic missile submarines, and aircraft supercarriers. They calculate that the benefit of effectively unlimited fuel supply is worth the cost for those ships. The submarines can stay submerged for months at a time and have extensive range, and the carriers can move through the ocean, displacing 100,000 tons of water, to anywhere on the globe without worrying about the logisitcs of the enormous amount of fossil fuel it would require.
Why do we have to be constrained by economics unlike the navy? The state within the state has access to nuclear powered ships so it is just a question of widening that capability to the members of the state living outside the walls of the keep. Not a matter of establishing precedent or anything. That part is done and long proven. The federal government even has experience managing water and power for entire geographical regions today.
I haven't seen any data that backs that up based on general principals. Most of the cost is in artificially-imposed operational requirements - however well founded.
Also, remember that nuclear, unlike solar, has a lot of room for improvement still, both in how it's done, and how it's regulated. Solar has already been tremendously optimized, while nuclear has not.
The cost argument seems to be advances by the same people who impose or support the additional operational requirements, and who also just have a philosophical aversion to nuclear power.
There are hundreds of nuclear power plants already in operation, many decades old. There have been only a very small number of minor accidents (3 come to mind: Chernobyl, Three mile island and Fukushima), in which only a few dozen people were killed. Nuclear, even using old technology, has proven to be far safer and better for the environment than any scalable alternative, including solar. New designs are even safer.
Calling Chernobyl a minor accident is insane. We are lucky it wasn’t worse but even then most of Europes forests are still polluted from the fallout. People directly killed during the incident is not a great indicator of incident severity when we're talking about environmental pollutants.
Nobody died from installing asbestos insulation yet here we are.
> Most of the cost is in artificially-imposed operational requirements
Indeed. Once there was a wonderfully efficient, economical nuclear reactor design, better thermal efficiency than PWRs, could be refueled during operation, considerably cheaper to build… However, nobody is THAT keen to build more Chernobyls.
(The RBMK design really was quite impressive, provided you weren’t too concerned about, well, safety.)
The economics of nuclear energy are difficult, today. So much of the cost is upfront that getting the investment is problematic; unless you have a guaranteed price per kWh, it really is a huge gamble.
All I see is people with an aversion to solar and wind, that champion nuclear for purely ideological reasons. The aversion seems to be mostly driven by the fact that solar and wind were first championed by eco hippies, and some people seem to find it hard to bear that the eco hippies were right in this case.
Nuclear has had tremendously more cumulative R+D spend than solar and wind. The notion that it's less optimized is absurd. And this is where your bias shows: we have empirically proven persistent scaling laws for solar and batteries. We also have seen nuclear become ever more expensive over time. Yet you claim that these trends will come to an end, and in the case of Nuclear will suddenly completely reverse themselves without any evidence.
To also bemoan the burdensome operational requirements while championing it's safety record is internally inconsistent.
And in the end no one has so far actually built a place where you could store the nuclear waste long term, and the costs of long term storage are not even fully factored into the costs of today's nuclear power plants.
"All I see is people with..." - Maybe some do, but not me. I think solar is great, and wind too. I have solar panels on my roof that cover most of my family's usage. Geo would be awesome, and hydro can be great.
I'm all for doing more, and improving our lot incrementally over time. Let's focus on doing more wherever we can.
Why is nuclear getting more expensive over time? Are we forgetting how to produce it or something? Actually we've been finding more efficient and safer ways to produce nuclear for decades, but we impose - as I said - artificial burdens that make it more expensive, or simply don't allow it at all. At least in the US.
The operational requirements DON'T make it more safe though, they just add cost. Storing nuclear waste is also safe, easy and cheap - if we allow it to be so.
Seriously, that's good to hear. I really often encounter arguments that seem heavily ideology driven.
Do you have some sources for operational requirements not making things safer? And as far as I know there are plenty of fusion concepts in the lab, but very few that have actually been explored at the full reactor scale. If you have any pointers on recent developments in that direction I would also be curious to take a look.
But then I also have to ask why nuclear? Why not methanation (or hydrogen if storage becomes feasible) and gas power plants or some more sophisticated version of that? That has much better complementarity to solar. And it typically is preferred to nuclear by energy system models wherever seasonality is strong.
Also I haven't really seen any proposals discussing long term waste storage. Again do you have any sources that discuss this?
Nuclear has had a century almost to hone its craft.
I'm a lftr fanboy, but nuclear had its time to optimize.
I can't believe you're calling Chernobyl "minor". Go take a vacation there if you disagree.
Anyway, nuclear is not cost competitive in the real world, imo it never will be with solid fuel, nor will it be safe. Certainly not with standards like yours.
Solve the waste problem and you solve nuclear. Waste is still the giant elephant in the room and a lot of people have a fifth grade solution to the problem (we will bury it under ground! We will fly it to the sun! We will resuse it until it is no longer radioactive!)
I used to have a neighbor who worked for the DOE, all of the viable solutions are blocked by people who don’t want it in their backyard. Can’t really move forward until that is solved..
> Waste is still the giant elephant in the room and a lot of people have a fifth grade solution to the problem (we will bury it under ground! We will fly it to the sun! We will resuse it until it is no longer radioactive!)
The correct solution: put it into dry casks and do nothing right now. Store it simple underground storage facilities or on the grounds of active nuclear power plants.
The casks are fine for the next 300 years, and during that time we can either:
1. Perfect the nuclear fusion, it will provide plenty of neutrons to transmute the waste.
2. Perfect fast fission reactors. See above.
3. Use some of the excess of too-cheap-to-meter green energy for accelerator-driven subcritical fission reactors.
4. Yep, use rockets to slowly launch the waste into space. We can already design a storage capsule that can survive re-entry.
In any case, we have literally hundreds of years to come up with a solution and there are many viable paths.
Yes. We're _already_ launching nuclear waste into space, in the form of RTGs with Pu-238. So there's been a lot of work towards making them passively safe, although the weight constraints for deep-space craft necessarily limit the amount of achievable safety.
> Even if a capsule could survive reentry, surely it wouldn't survive impact.
It'll be moving at a terminal velocity, and can be engineered to not fragment on impact.
For example, I remember reading about a proposal to alloy the nuclear waste elements with a carrier metal like iron or nickel, and then cover them in an ablative graphite shell.
>Waste is still the giant elephant in the room and a lot of people have a fifth grade solution to the problem (we will bury it under ground! We will fly it to the sun! We will resuse it until it is no longer radioactive!)
Could the waste be 'sent' into space?
Bonus points for sending a certain human with it.
Serious question, though.
This was one of the Soviet Union’s proposed use for Energia (a super-heavy launcher which flew precisely twice before the Soviet Union collapsed). In practice, there would be, ah, challenges; no launcher ever built is reliable enough that anyone would be particularly comfortable with _launching large amounts of high-level nuclear waste_ with it.
> Could the waste be 'sent' into space? Bonus points for sending a certain human with it. Serious question, though.
Would you want a RUD of nuclear waste in the atmosphere? That's the key thing with sending stuff to space, we are nowhere near close enough in terms of reliability and cost to what would be needed to send the stuff away.
By "space" do you mean low earth orbit - where the stuff will reenter the atmosphere within (say) a century? Or geosync orbit - where it'll stay up there forever-ish...but ain't actually gone? Or actually gone, like (say) Mars?
IIR, the current rock-bottom (Falcon 9) launch prices are something like $1,000/lbs. to low earth orbit, $2,500/lbs. to geosync, and $6,500/lbs. to Venus.
A quick Google says the US has about 88,000 tons of radioactive waste. So - 88,000 tons = 176,000,000 lbs. = $176,000,000,000 just to put it in low earth orbit. And something like 4,600 Falcon 9 launches. (Some fraction of which would doubtless go badly wrong, spreading radioactive stuff all over the landscape.)
In short - it's a cool-sounding idea. But neither the numbers nor the politics are remotely near viable.
Lftr is supposed to be in the high nineties allegedly of fuel consumption %.
And the gas/liquid nature means separation and processing is a lot more viable
Contrast that with solid fuel rods using like 20% of its fuel, and that only of the fissile uranium isotopes. The the waste is locked in a solid form of the spent fuel rods
I'll take lftr which (theoretically) is so much cleaner.
Fusion isnt without waste, the high speed neutrons irradiate/transmute the reactor
Burying the waste is an excellent solution. We are currently dealing with the much more dangerous problem of hydrocarbon combustion waste by releasing it into the atmosphere.
What problem with the waste? Reactors are working and generating power today for many thousands of americans. Whatever is being done with waste today seems to work well enough to continue reactor operations without any major headlines. Just seems to be a bit of cognitive dissonance here between what is claimed online and what we see today out in the field generating power.
Current waste takes tens of thousands of years to stop being radioactive. Until then we need to hide it from all living life. In the meantime we continue to create waste.
We are getting better shrinking from thousands to a single thousand.. but we still create waste while we argue where to store the new waste.
Many countries would have to buy nuclear fuel from other countries much like they do for gas and oil. On the other side very few countries can build their own solar panels so it seems the same sort of problem. However if you accept to depend from potential hostile countries at least solar panels don't do much harm when they fail and it takes very little effort to install the equivalent amount of power of a nuclear power plant. If we only could all get along and have a global power grid with always 12 hours of sunshine on it.
Until we solve the storage problem, the manufacture of batteries required to store solar power at night and during other low generation times is actively doing harm even without failing.
I don’t think this invalidates your point but I do think it is incredibly important to recognize that environmental harm done slowly over time is no less impactful than that done by a disaster.
Unfortunately the arrangement of the continents would require undersea transmission lines for this. Would be an interesting future where the Bering Strait is the most valuable real estate on Earth for the American-Euraisian grid connection. But more likely we would build fusion reactors or get over our nuclear phobia before building something like that.
Most country that can invest a few hundred million can build their own solar panels that will be significantly cheaper than existing and future fossil fuel prices. It's just those panels will be 2x+ the price of those produced in China. So can't compete in the free market.
This paper just reads like an attempt at sounding smart while actually saying little.