I've always been curious why a cost-effective widespread implementation of geothermal energy has never been considered a holy grail of energy production, at least not in the public debate. Much of the discussion is so focussed on nuclear fusion, which seems so much harder and less likely to be reliable.
Since you're comparing it to nuclear, I'm assuming you mean electricity production here, not energy production?
It's always worth remembering that electricity only accounts for ~20% of global energy consumption (in the US it's closer to 33%).
I suspect people confuse these two because in a residential context electricity plays a huge part of our energy usage, but as a whole it's a smaller part of total energy usage than most people imagine.
But any serious discussion of renewable energy should be careful not to make this very significant error.
At home I use 15,000kWh of oil for heating each year (about 10kWh per litre, 1500 litres), and 8,000kWh of electricity (we use a lot more than the average household). For driving that's another 5000kWh a year if at 4 miles per kWh.
So even in a residential context, electricity is only about 1/4 of the demand. Across the whole country it's less than 300TWh out of 1500TWh, under 20%.
That excludes "imported energy" though, as in goods which used energy to make but were then imported.
Driving can push up the low points (charge cars overnight), but heating would put a lot of demand in winter months, meaning a day time cold day in January with no wind will require a lot of dispatchable electricity, at night time in September with a gale blowing wind will be providing almost all the demand.
Nuclear doesn't really help as it's more expensive than the wind when it's windy and demand is low, and its impossible to build enough to cover the peak January demand unless you spread the fixed cost over the entire year, which means getting rid of every other form of electric production, and you'd still end up paying more per kWh than you would with other forms of storage.
Nuclear can't survive in a free market. It can't scale up to provide for areas of high demand, low supply, and it can't scale down to be affordable when there's high supply and low demand.
Only about 30% of the energy in gasoline is converted to useful work in a gasoline car (the 'make metal box go forward' part). The remaining 70% is Rejected Energy (the steam you see going out the tail pipe in winter).
Which (not sure if you did this intentionally or accidentally) brings up an interesting point on the parent comment and the LLNL sankey:
> It's always worth remembering that electricity only accounts for ~20% of global energy consumption (in the US it's closer to 33%).
That "global energy consumption" figure includes a lot of Rejected Energy going out tailpipes and smoke stacks turning burnables into electricity. A secret bonus of wind and solar is if you produce electricity without burning things, you actually decrease the energy demand! If you're not losing 70% of your energy consumption to the Rejected category, you suddenly need a lot less total energy.
Rejected energy means energy that is lost as waste heat without performing any work first. For example, a coal fired power plant may generate 3 megajoules of thermal energy from coal combustion but only deliver 1 megajoule of it as electricity. The other 2 megajoules are lost as useless waste heat.
The 1 megajoule of useful electricity is also ultimately dissipated as low grade heat, but it can do work first (like generating light, or pumping water uphill).
I live in a part of the world that is far below freezing for a significant portion of the year. Thus a large portion of my annual energy usage goes into not freezing to death.
When I drive my daughter to school when it’s -40 fucking degrees, a lot of the energy I use goes into heating my vehicle, swearing, moving and swearing. But this energy also leaks through my windshield, through my exhaust system and through my engine. This energy (heat) doesn’t provide any benefit to anyone and just leaks out into the atmosphere (which we’ve already established is trying to kill me).
That’s rejected energy. Or when it’s below -40, rejected motherfucking energy. :)
A IC car’s heating system normally taps into the engine’s cooling system, so that heat is mostly free. In a pinch you can actually turn the heater on full to help cool the radiator.
I had to do that when my radiator sprang a leak on the freeway and the engine heat kept creeping up. Unfortunately it was late summer and not at all pleasant.
I managed to get to a gas station with some stop leak in stock... If they didn't, I was ready to crack an egg in it.
I once rode in a friend's car in a similar situation. Very much not pleasant. His problem was a thermostat problem, so he hadn't lost all cooling but enough we used the same workaround. Running the heat in the summer time resulted in a couple of very sweaty dudes.
Drilling is one of those things which used to be extremely expensive but has very gradually come down in price. Thanks, ironically, to the oil industry. It's unsexy because there's no "silver bullet" waiting in the wings.
It's also quite hard to find suitably hot rocks suitably close to the surface.
Focusing on fusion .. I think that's a legacy of 60s SF, when the fission revolution was still promising "energy too cheap to meter".
> Drilling is one of those things which used to be extremely expensive but has very gradually come down in price. Thanks, ironically, to the oil industry. It's unsexy because there's no "silver bullet" waiting in the wings. It's also quite hard to find suitably hot rocks suitably close to the surface.
That's basically it. Most geothermal plants today are in locations where there are hot rocks, maybe geysers, close to the surface. "Deep geothermal" gets talked about,
because temperatures high enough for steam are available almost everywhere if you can drill 3,000 meters down. There are very few wells in the world that deep, not counting horizontal drilling runs.
The economics are iffy. You drill one of the most expensive wells ever drilled, and you get a medium-pressure steam line. Average output is tens of megawatts.[1]
The economics change when you’re in oil country. My beautiful little province has oil wells drilled between 250 and 2900 metres. Due to corporate ‘issues’ many of these wells are orphaned and remediation becomes a provincial problem. With deep holes and provincially owned electricity and gas companies, geothermal makes more economic sense; it’s robbing a benefit from a big cost centre.
I went to high school with two guys who are working on geothermal as a means to remediate orphan wells. I’m biased in their favour, but the numbers make a lot of sense.
To be fair, that promise of fission made sense from a purely scientific and mathematical perspective, before running into the practical realities of how its externalities interact with real-world politics. Fission is expensive because in practice it turns out we care quite a lot about proper waste management, non-proliferation, and meltdown prevention.
In a world where anyone could just YOLO any reactor into production with minimal red tape, consequences be damned, fission energy would actually be extremely cheap. Hence the optimism around fusion. The promise of fusion is an actualization of last century's idealistic conception of fission. It can be a silver bullet for all intents and purposes, at least once it's established with a mature supply chain.
I fully understand that waste management of fission reactors is a Very Big Deal. But I still stand behind the argument that opposing nuclear power in the 70s and onward is possibly the biggest own goal the environmental movement has ever achieved.
At worst, nuclear waste contaminates a discrete section of the Earth. Climate change affects literally everywhere. The correct answer would have been to aggressively roll out fission power 40-50 years ago and then pursue renewables. You can argue that other solutions would make fission power obsolete, but we would have been in a much better spot if it'd at least been a stepping stone off fossil fuels. Instead, we have 40-50 years of shrieking and FUD from environmentalists over an issue that can be kept under control with proper regulation. The US Navy has operated reactors for over 60 years without incident, proving it can be done with proper oversight.
TL;DR nuclear has issues, but I'd take it over coal every day and twice on Sundays, at least until something better can scale.
People want houses. Planners can either yell and stomp their feet about this or adapt to circumstances. It's like electric cars. People want cars. Better they have the ability to have an electric SUV or pickup, because if you try to force them into little tiny econoboxes or lecture them about how they should really be using mass transit, they're just going to flip you the bird and walk away.
Similarly, better to have people be able to have reasonably energy-efficient houses than demanding they all live in apartments.
People want a place to call home. Those come in many shapes and sizes. Denser living does not mean a smaller living space. By building 'up', you can provide both.
The only ones demanding anything are those who show up to try and stop apartments.
Its funny how mass transit was once seen as a way out of the crisis of too many city horse carriages and all the manure they produce. Metro and rail was simply considered the natural solution, which combined with denser living space allowed for labor intensive industries to sprout up around cities. But then came the horseless carriage and suburbian sprawl became a thing and now we have too much of that. Too much hardened soil and flooding is becoming a much bigger problem, animal and insect populations wither, road networks have upkeep, transport and storage costs make everything imperceptably more expensive. In my scifi fantasy future vertical farming will become a thing which should allow a major shrinking of hinterlands around dense population areas, though preferably not replaced with suburbia. Maybe we could go live in the clouds too.
(Source needed. This probably depends on a lot of variables in play.)
Plenty of people in dense urban areas are happy with living in an apartment and, where I live, buying a condo in the city is at least as frequent as buying a house 20 km away from it for the same price.
Living in suburbia has its downsides - long commute, very limited entertainment and cultural possibilities, very limited choice in schools. Not everyone loves cutting the lawn etc. either, I surely don't. If any of your family members has any disease that could flare up, ambulance response time tends to grow worse with the growing distance.
Of course, a lot depends on factors such as "is the transport authority willing to make public transport actually safe and nice". That requires keeping raving drugged lunatics out of it, plus paying enough money for it. AFAIK in the US, Republicans have an ideological problem with the "paying money for it" part and the Democrats have an ideological problem with the "suppressing antisocial behavior in it" part.
People want a lot of things, many of them conflicting. I'd love a huge house on a large lot in a walkable area and it to be cheap, and also close to nature. Letting markets work is a good way of resolving people's revealed preferences. Some will prefer a condo in a walkable area, others a large lot outside a less expensive city, others will pay through the nose to have a single detached unit in a high cost of living area.
The problem with that argument is that nowhere did environmentalists in the 70s or 80s prevent nuclear power plants from being build. Nuclear has received much more subsidies than solar or wind ever did (even if we ignore the indirect military->civil subsidies) and it still never became economical. Back in the 70s and 80s coal was much cheaper, and now solar and wind are so much cheaper it doesn't make any sense to invest in nuclear. The nuclear power plants that were build, got built largely for political reasons (energy independence, and military), and the reason why not more got build was not those pesky environmentalists, but that it was expensive.
I’m relatively sure that at least in Germany the environmental movement had a lot of influence preventing the aggressive pursuit of nuclear power. Your point about subsidies still stands of course, but economics of power generation notoriously ignore the costs of climate change.
I agree. I think the correct environmentalist position at that time wouldn't have been to oppose nuclear, but to advocate for improvements, streamlined approvals of improved designs, and public investment or incentives.
I wasn't really commenting on the merits of 20th century environmentalist movements, more raising the general point that fission power has inherent costs which weren't reflected by narrow 1950s analyses of how much energy was extractable from U-235. Operation of a fission plant requires much more capex and opex than it would if we didn't care about cleanliness (waste management), security (fissile material theft prevention), or safety (meltdown prevention).
Fusion power is more complex to invent and practically depends on modern technologies that didn't exist 50 years ago, but once the first demonstration plants are operational, marginal costs to deploy and operate more should be much lower and ultimately become very low at scale.
Back then, it affected everyone in two ways, which were the things Greenpeace campaigned against: nuclear weapons, especially overland testing, and dumping waste at sea.
Chernobyl took out Welsh farming for years, and in a few places decades, because it spread a thin layer of bioaccumulative poison over the whole of Europe.
Neither of these have anything to do with running a well-regulated nuclear power program. Chernobyl happened because of the apathy and incompetence endemic to any Marxist-Leninist system, not because a modern democratic state is incapable of regulating the nuclear power industry.
Know what else spreads a thin layer of poison over the whole of the world? Coal power.
Democracy just as lazy and apathetic is whatever the USSR counts as; the point of capitalism (which is different to democracy) redirect the laziness into something more productive — this works to an extent, but depends on competition which is greatly reduced in the case of nuclear reactors.
That it's a different axis than democracy-communism is also why the not-at-all-democratic military reactors around the world seem to be doing fine.
Many others here have talked about the difficulties of geothermal, which doesn't really get to the heart of my question: why the lack of hype around breaking down those difficulties? I appreciate that you took the time to comment on why it isn't so sexy, the SF argument probably has a lot to do with it.
The problems are that rock isn't a good conductor of heat, so once you've cooled a bit down, you have to wait for it to warm up. Warming only happens very slowly at the rate of < 50mW / m² which limits the amount of power you can get out.
Probably because not everywhere on earth has the same easy access that Iceland has. The article mentions this:
> There aren’t gates of Hell just anywhere. A kilometre below ground in Kamchatka is considerably hotter than a kilometre below ground in Kansas. There is also readily accessible geothermal energy in Kenya (where it provides almost fifty per cent of the country’s energy), New Zealand (about twenty per cent), and the Philippines (about fifteen per cent)—all volcanic areas along tectonic rifts. But in less Hadean landscapes the costs and uncertainties of drilling deep in search of sufficient heat have curtailed development.
Until recently, the geographical locations where geothermal is feasible and economic was very limited. Ironically it is tech from fracking/shale gas that is starting to open up a far wider range of possible sites at lower cost.
Because unless you sit on top of a volcano, amount of renewable geothermal energy is minuscule. In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!). In other words, in most places geothermal plant acts more like a limited battery powered by hot rock, so unless drilling is extremely cheap, it does not make economic sense compared to other energy sources.
> In most places on Earth it's somewhere around 40 mW/m2 (i.e. accounting for conversion losses you need to capture heat from ~500 m2 to renewably power one LED light bulb!)
Ground-source heat pumps extract about 1000 times more power from ground loops, where does the difference come from?
A number of sources. Often the air above - ground source relies on the ground being the average temperature of the year round air once you get deep. They also tend to run in heating mode half the year, and cooling mode the other half.
Ground-source heat pumps are irrelevant to geothermal energy sources, and it's unfortunate that the article mentioned them. Ground-source heat pumps are just storing heat from the air during the summer and retrieving it during the winter.
Hmm, that doesn't sound right. Many homes don't use these heat pumps for cooling in the summer because getting convectors and plumbing installed is just as expensive as getting a separate AC, and then you have some redundancy in the system too.
I'm sure that battery effect is a factor but it must be a relatively minor one.
It's not minor at all. For a ground-source heat pump to work at all, you need to sink the pipes deep enough that the soil temperature is pretty stable year-round, which means the heat it can exchange with the aboveground air is minor or insignificant. So the vast majority of heat flux into or out of that soil is due to the heat exchange fluid circulating in the pipes.
After a few years of pumping heat out of the ground below the frost line during the winter, they'll freeze the ground solid and stop working (and possibly destroy the foundation of the house in the process, since often the pipes are installed in trenches around the house).
The only exception is if they're one of these few borderline systems that drill so deep they really are bringing up fresh energy from the depths, like some of the systems mentioned in the article.
Insulation, adobe construction, and vigorous exercise can all "heat houses in winter" in the same way as ground-source heat pumps, but none of them can be meaningfully compared with hydroelectric or nuclear power except in a specific situation. How much insulation is enough to charge your cellphone or run a load of laundry? The question is nonsense.
Ground-source heat pumps (with a few exceptions and borderline cases like those mentioned in the article) are not sources of heat, so you can only "get heat from" them in the short term; as with a battery or an interest-free checking account, though you may be able to temporarily run a debit balance, in the long run you can only get out what you put in.
This is a fundamental difference from energy sources.
Energy storage is an important complement to energy sources, especially renewables, and can substitute for energy sources to a limited extent, but confusing them is a fatal error.
Ground source heat pumps as used for heating buildings source their energy from the sun and work very well. They are outcompeted by air source heat pumps despite better efficiency because digging is expensive.
As I understand it, no, they do not source their energy from the sun. You have to bury the pipes below the frost line. In temperate climates, the frost line is the place where the earth is so well insulated from the surface by the thickness of earth above it that heat conduction is insufficient to freeze it throughout the entire winter. The sun only heats that surface by radiation.
> Unlike air-source heat pumps that struggle to extract heat from frigid air, ground source heat pumps tap into a remarkably stable heat source: the earth itself. Below the frost line, ground temperatures remain relatively constant year-round, typically between 50-60°F (10-15°C). This consistency makes ground source heat pumps highly efficient even during the coldest months.
If you were building a ground-source heat pump to heat your house in the summer, you could get away with burying the pipes at a much shallower depth and in effect converting the earth into a low-temperature passive solar collector. But generally people want to heat their houses in the winter instead.
While it's true that a geothermal plant is a limited battery powered by hot rock, that doesn't mean it doesn't make economic sense compared to other energy sources.
The worst earthquake that was induced that way was 3.5, but given that one of the quakes happened in an area that had a catastrophic earthquake in the Middle Ages, some caution might be warranted: https://en.wikipedia.org/wiki/1356_Basel_earthquake
The reactor breaking and taking a very long time to repair because the repairs would have to be done remotely, with robots. The structure becomes too activated for people to go inside, even after the reactor is shut off.
The reactor breaks because it's a large device operated at high stresses (power/area, neutron loading). There are many components and joints that can fail.
BTW, this means fusion will be expensive, because getting all those components to be reliable right off the bat becomes expensive. No tiny cracks in the welds means expensive quality control.
all the machinery used to obtain and maintain an economically viable fusion reaction. Having worked with particle accelerators and synchrotron rings, I'll tell you that stuff breaks down all the time.
I think it mainly depends on how easy it is to access that energy. I went to Tuscany last year and to my surprise there were geothermal plants everywhere. I have never heard about these plants beforehand, but here they are in Italy quietly powering the countryside and heating greenhouses to grow basil all year around.
There is a crazy amount of energy available everywhere but it is not in the interest of the very powerful very wealthy existing players. This isn't some grand CONSPIRACY. For example oil companies may construct energy investment portfolios that would quite sensibly acquire promising energy related research. They do a simple cost benefit analysis then chose to modestly further research it or shelve it. They turn it into valuable pieces of paper that accumulate value over time. What is there for them not to like about it?
I like how David Hamel put it: We live in this thin sliver on the surface of the planet where it is reasonably peaceful. This is the tranquility! It's a good thing! If you go up or down by a mere few miles there is so much energy it kills you.
Well, the survey is constitutionally mandated, whereas the ACS is just a law, so they exist on two different levels of enforceability and responsibility. The government can't not do a census; it's not permitted. The ACS can have its rules revised with no problem.
> A core value of America is our right to obstruct any government attempt to improve our lives and I defend that stubborness.
I think this is a reinvention of history, because much of American history, and the writings of the founders, do not seem to imply this. The core value to my understanding is "no taxation without representation", probably followed by freedom of speech (from government). I don't think this is true anymore though, given how many people are happy for the king to impose taxes on them at will.
The author is first discussing the importance of orthonormal bases for function approximation. This is a crucial first point of understanding from the mathematical point of view. Then comes periodicity as a key criterion, then the spectral theorem... From the mathematical point of view, the FFT and DFT is really one of the last things that should be examined.
Judging from the past, making the census voluntary in the same way as the vote in federal elections would have this peculiar capacity to be used to skew the results in favor of any result the designers should choose. So of course this should be pushed by the Republicans. It's much easier to avoid unpleasant facts when you can effectively erase the existence of inconvenient groups of people.
Although that is a possible reason the stated reason for more than a decade is minimal constitutionality. Republicans and conservative groups have long claimed to advocate for smaller government and minimal access to any data except for that required to perform government's mandates according to the various governmental agency charters.
The reason for the census comes directly from the US constitution and its only stated purpose there is to redistribute congressional districts and Electoral College electors.
My own personal opinion favors giving government as much access to data as possible because contrary to what many people claim government is overwhelmingly more productive compared to the private sector.
> contrary to what many people claim government is overwhelmingly more productive compared to the private sector
People always make this about public vs private sector but in my experience, it has more to do with the size of the organization. Large private sector organizations are just as susceptible to the slow-as-molasses bureaucratic processes as big government. Similarly, I have seen local governments be as fast-moving and agile as a startup. The simple reality is, the more people and processes are involved, the longer things take.
> to what many people claim government is overwhelmingly more productive compared to the private sector
Whoa, that's a very strong statement that requires some refinement I think.
In any case, I understand the claimed reasons, but I remain skeptical. Sometimes making something "voluntary" is not in the interests of freedom or small government. I'm sure the founding fathers were aware of that.
> to redistribute congressional districts and Electoral College electors
That's extremely important and has been used to "remarkable" effect in recent years.
> that's a very strong statement that requires some refinement
Yes, but much of that refinement would be the gritty details of pushing back on awful self-serving definitions that were carefully crafted to mislead. Flouting them altogether is a strong opener.
Contrast to the boring analytical speech: "The notion of value espoused by neoliberal economics is wealth-weighted while the colloquial definition of the word does not have a wealth-weight attached, sometimes even the opposite (see: feeding orphans). This loophole is large enough to march 1000 elephants through and wage a class war. "Value Creation" is not about doing what people want, it's about doing what wealth-weighted people want, and as inequality grows that increasingly means doing what rich people want, which is primarily to pump assets so that they can get paid for being rich. This twist of terminology is how you can brainwash someone into thinking that enshittification, in all its forms, is somehow for the greater good, when it's actually just for the good of rich people who want to get paid for being rich."
The boring analytical speech is theoretically the stronger argument, but if theoretically stronger arguments won elections we wouldn't be here. So the best move is just to reverse-uno the "government bad, drown it in a bathtub, private sector good" propaganda.
But even the analytical speech is too strong: humanity has been down that road and it is fraught with other problems. Government isn't necessarily better either and like everything it must be assessed on a case-by-case basis. Enshittification is inevitable, but would be easily undone if the anti-trust laws were properly enforced. It's a beautiful balancing act, and we can't get there by black-and-white "government bad/good, private bad/good". At least, that's what I think.
It seems that people will only willingly act in the common good for small communities; at the level of government, you either enforce the common good, or you take advantage of greed and try to loosely direct it into the interests of the common good. Right now there is an argument to be made that we are not successfully achieving the latter strategy as "Value Creation" is now a bastardization of its original intent. But the former option is too diabolical to consider.
When you drive to work, you yearn for a negotiation with the local road monopoly?
On cold nights, you dream of a glorious future where an unregulated energy utility could try to trick you into surge pricing?
When you wrote your post, you did so with teeth clenched in disgust at having to use a government invention?
When you take the trash out, you wish it were a bit more exciting, that you had to dodge gunfire from skirmishing warlords?
And so on, and so on, but the point is that a statement you probably intended as a slight hyperbole actually required a staggering amount of "out of sight, out of mind."
What a bizarre argument. Where is the implication even coming from that gets you to the idea they can't think government services can be better for the people than private products unless you "agree with the US's current budget". Why would the current us budget, good or bad, efficiently spent or not, deeply corrupt or perfectly honest, be the only criteria for having an opinion on shared services versus private market?
Are you trying to say that you think the current US budget is bad and therefore all government spending for all time is inherently bad and worse than every other option? If so that's a really weak take.
It depends. The military is not particularly efficient, but it's not clear whether that is desirable. The tax system is not efficient, but that is by design of the industry. Many other organizations are surprisingly efficient actually, hence why some of DOGE's scalpel treatment was so harmful.
The government is an enormous ship and you want enough checks and balances to ensure it cannot suddenly turn on a dime without a huge emergency.
As for the debt, that mostly comes down to:
1) Military spending to enforce US domination across the globe.
2) Social security for the Boomers (many countries are being hit by this).
3) Unsustainably low tax rates given the requirements of government. In particular, the presence of numerous tax loopholes for the ultra-wealthy.
4) The insanity of the healthcare industry in the US.
The meaning of the word "efficiency" is highly contextual.
Operation Warp Speed rapidly developed and distributed a totally novel COVID vaccine to billions of people in just two years, but was extremely costly. Was that efficient?
The number of federal employees per American citizen has been shrinking for 80 years. Is that efficient, if it also leads to long waits for government visits?
The United States has not fought a peer state in war for decades - largely because billions of dollars are spent to build bombs that will spend their entire lives in a warehouse. Is that efficient?
Large organizations tend to operate the same way regardless of their ideology. The only real difference between government and private orgs is that one of them will fail faster under stress - and much like efficiency, whether that's good or bad depends on context.
When "higher turnout" is still low enough in national elections that the missing amount is more than the differential in the popular vote, this doesn't disprove anything. You're citing an approximate statistical poll to try to mitigate that, but the entire point of the census is that right now the way it works is to try its best _not_ use those methods and instead attempt to actually get responses from everyone.
"Modeled" is the key point here. If one accepts the premise that modeling is close enough to reality that the differences aren't meaningful to be worth measuring with a full census count, then the change to census collection won't seem problematic. If one rejects that premise, it follows that the difference between the statistical modeling and the reality of a full census count might be significant enough to be worth measuring. That "difference" is a chunk of people who might not fit the expectations of the modeling, and it would be entirely expected that the ones who would benefit more from the model's expectations compared to the full census count would be the ones who are in favor of using the modeling.
> That's incorrect. Since the Trump era, it's Republicans that benefit from higher turnout, while Democrats benefit from low turnout
The variance in voting preferences dominates this effect. We simply have insufficient data to make statements like this confidently, a fact being laid bare by the potential fuckup that has been the Texas redirecting.
Both Pew and Blue Rose Research have confidently stated that Trump would have won by 3-5 points if everyone had voted. These conclusions were based on thousands to millions of samples, far more than a typical poll.
> Both Pew and Blue Rose Research have confidently stated that Trump would have won by 3-5 points if everyone had voted
Correct. Re-run that analysis today, however, and the benefit of turnout may have shifted. That’s the point. There isn’t a consistent partisan tilt in turnout.
In 2020, turnout did not benefit Trump. I suspect, given his economic track record, it will not again in 2026 and 2028.
It may be clear with the benefit of hindsight that a higher turnout in 2024 would have further benefited Trump. But the Republicans have not typically acted with the intention to make it easier for arbitrary members of the public to vote (happy to be corrected), and I don't expect them to act in this interest in the future.
Republicans and Trump supporters are two overlapping but distinct groups. A lot of establishment Republican approaches were based on ideas that are obsolete in the Trump era. Trump can win enough Bangladeshi immigrants in Queens or Hindi nationalists in New Jersey to win the popular vote. Mitt Romney Republicans can’t.
There is nothing about right-wing politics that inherently benefits from a limited electorate. After all, Latin American countries regularly flip-flop between far-left and far-right leaders, like in Brazil with Bolsonaro, or Modi in India. The increase in immigrants in the U.S. and expansion of the electorate to include them won’t get rid of “Republicans,” like many had hoped. Instead, it simply changed what kind of right-wing party could form a viable majority coalition.
At the polls, Republicans primarily benefit from Republicans not being in office. Being the opposition party lets them trick more people with their signature fake appeals to liberty and patriotism like small government, fiscal responsibility, the second amendment, anti- war and foreign interventions, international strength, claims of political persecution, demands for accountability, etc.
When they're actually in office it's obvious that they have absolutely zero intention of following these ideals, but rather just abuse them in hypocritical displays of naked power. So it's only a diminishing number of true-believer type followers that continue twisting themselves into every-tighter knots of cognitive dissonance.
This is why they often start flashy wars - to distract from outright harmful domestic agendas and try to bolster support based on some common external enemy. Tramp himself is of course also trying to distract from the revelations of long term spicy kompromat hanging over his head. The things some people will do to get a loan, smh.
Seeing every issue in black and white is simplistic, and can cause issues. Seeing shades of gray is better. Attempting to avoid bias by seeing only a single shade of gray, as you are suggesting, is obviously not going to be very effective for percieving the world.
"Language in a pending 2026 spending bill written by the Republican-led U.S. House of Representatives goes further. It would make both the ACS and the regular 10-year census voluntary, and would also prohibit the agency from reaching out more than once to anyone who doesn’t initially respond."
How many decades did it take for neural nets to take off?
The reason we're even talking about LeCun today is because he was early in seeing the promise of neural nets and stuck with it through the whole AI winter when most people thought it was a waste of time.
But neural nets were always popular, they just went through phases of hype depending on the capacity of hardware at the time. The only limitation of neural nets at the time was computational power to scale up. AI winters came when other techniques became available that required less compute. Once GPGPU became available, all of that work became immediately viable.
No similar limitations exist today for JEPA, to my knowledge.
Depends on how far back you are going. There was the whole 1969 Minsky Perceptron flap where he said ANNs (i.e Perceptrons) were useless because they can't learn XOR (and no-one at the time knew how to train multi-layer ANNs), which stiffled ANN research and funding for a while. It would then be almost 20 years until the 1986 PDP handbook published LeCun and Hinton's rediscovery of backpropagation as a way to train multi-layer ANNs thereby making them practical.
The JEPA parallel is just that it's not a popular/mainstream approach (at least in terms of well funded research), but may eventually win out over LLMs in the long term. Modern GPUs provide plenty of power for almost any artifical brain type approach, but of course are expensive at scale, so lack of funding can be a barrier in of itself.
Consistency is one aspect, but it is not enough. I believe (and this is somewhat based in other arguments from neuroscience and discussions with alignment researchers) that two more are necessary: compression, which demonstrates algorithmic development; and linear representation capacity, as this is the only way that we really interpret the world, and therefore will only define another as intelligent if it can distill knowledge into the same language that we understand.
I think compression is probably a natural consequence of coherent self models? Isn't requiring other minds to package their intelligence in human interpretable linear narratives is like requiring dolphins to demonstrate intelligence through written language?
> So the question is, how intelligent are LLMs when you reduce their training data and training? Since they rapidly devolve into nonsense, the answer must be that they have no internal intelligence
This would be the equivalent of removing all senses of a human from birth and expecting them to somehow learn things. They will not. Therefore humans are not intelligent?
> LLMs dont do that; they just lack understanding.
You have no idea what they are doing. Since they are smaller than the dataset, they must have learned an internal algorithm. This algorithm is drawing patterns from somewhere - those are its internal, incorrect assumptions. It does not operate in the same way that a human does, but it seems ridiculous to say that it lacks intelligence because of that.
It sounds like you've reached a conclusion, that LLMs cannot be intelligent because they have said really weird things before, and are trying to justify it in reverse. Sure, it may not have grasped that particular thing. But are you suggesting that you've never met a human that is feigning understanding in a particular topic say some really weird things akin to an LLM? I'm an educator, and I have heard the strangest things that I just cannot comprehend no matter how much I dig. It really feels like shifting goalposts. We need to do better than that.
In split-brain experiments this is exactly how one half of the brain retroactively justifies the action of the other half. Maybe it is the case in LLMs that an overpowered latent feature sets the overall direction of the "thought" and then inference just has to make the best of it.
Yeah, that's compression. Although your later comments neglect the many years of physical experience that humans have as well as the billions of years of evolution.
And yes, by this definition, LLMs pass with flying colours.
I hate when people bring up this “billions of years of evolution” idea. It’s completely wrong and deluded in my opinion.
Firstly humans have not been evolving for “billions” of years.
Homo sapiens have been around for maybe 300’000 years, and the “homo” genus has been 2/3 million years. Before that we were chimps etc and that’s 6/7 million years ago.
If you want to look at the entire brain development, ie from mouse like creatures through to apes and then humans that’s 200M years.
If you want to think about generations it’s only 50/75M generations, ie “training loops”.
That’s really not very many.
Also the bigger point is this, for 99.9999% of that time we had no writing, or any kind of complex thinking required.
So our ability to reason about maths, writing, science etc is only in the last 2000-2500 years! Ie only roughly 200 or so generations.
Our brain was not “evolved” to do science, maths etc.
Most of evolution was us running around just killing stuff and eating and having sex. It’s only a tiny tiny amount of time that we’ve been working on maths, science, literature, philosophy.
So actually, these models have a massive, massive amount of training more than humans had to do roughly the same thing but using insane amounts of computing power and energy.
Our brains were evolved for a completely different world and environment and daily life that the life we lead now.
So yes, LLMs are good, but they have been exposed to more data and training time than any human could have unless we lived for 100000 years and still perform worse than we do in most problems!
Okay, fine, let's remove the evolution part. We still have an incredible amount of our lifetime spent visualising the world and coming to conclusions about the patterns within. Our analogies are often physical and we draw insights from that. To say that humans only draw their information from textbooks is foolhardy; at the very least, you have to agree there is much more.
I realise upon reading the OP's comment again that they may have been referring to "extrapolation", which is hugely problematic from the statistical viewpoint when you actually try to break things down.
My argument for compression asserts that LLMs see a lot of knowledge, but are actually quite small themselves. To output a vast amount of information in such a small space requires a large amount of pattern matching and underlying learned algorithms. I was arguing that humans are actually incredible compressors because we have many years of history in our composition. It's a moot point though, because it is the ratio of output to capacity that matters.
They can't learn iterative algorithms if they cannot execute loops. And blurting out an output which we then feed back in does not count as a loop. That's a separate invocation with fresh inputs, as far as the system is concerned.
They can attempt to mimic the results for small instances of the problem, where there are a lot of worked examples in the dataset, but they will never ever be able to generalize and actually give the correct output for arbitrary sized instances of the problem. Not with current architectures. Some algorithms simply can't be expressed as a fixed-size matrix multiplication.
>Most of evolution was us running around just killing stuff and eating and having sex.
Tell Boston Dynamics how to do that.
Mice inherited brain from their ancestors. You might think you don't need a working brain to reason about math, but that's because you don't know how thinking works, it's argument from ignorance.
People argue that humans have had the equivalent of training a frontier LLM for billions of years.
But training a frontier LLM involves taking multiple petabytes of data, effectively all of recorded human knowledge and experience, every book ever written, every scientific publication ever written, all of known maths, science, encylopedias, podcasts, etc. And then training that for millions of years worth of GPU-core time.
You cannot possibly equate human evolution with LLM training, it's ridiculous.
Our "training" time didn't involve any books, maths, science, reading, 99.9999% of our time was just in the physical world. So you can quite rationally argue that our brains ability to learn without training is radically better and more efficient that the training we do for LLMs.
Us running around in the jungle wasn't training our brain to write poetry or compose music.
No, I think what he said was true. Human brains have something about them that allow for the invention of poetry or music. It wasn't something learned through prior experience and observation because there aren't any poems in the wild. You might argue there's something akin to music, but human music goes far beyond anything in nature.
We have an intrinsic (and strange) reward system for creating new things, and it's totally awesome. LLMs only started to become somewhat useful once researchers tried to tap in to that innate reward system and create proxies for it. We definitely have not succeeded in creating a perfect mimicry of that system though, as any alignment researcher would no doubt tell you.
So you're arguing that "running around in the jungle" is equivalent to feeding the entirety of human knowledge in LLM training?
Are you suggesting that somehow there were books in the jungle, or perhaps boardgames? Perhaps there was a computer lab in the jungle?
Were apes learning to conjugate verbs while munching on bananas?
I don't think I'm suggesting anything crazy here... I think people who say LLM training is equivalent to "billions of years of evolution" need to justify that argument far more than I need to justify that running around in the jungle is equivalent to mass processing petabytes of highly rich and complex dense and VARIED information.
One year of running around in the same patch of jungle, eating the same fruit, killing the same insects, and having sex with the same old group of monkeys isn't going to be equal to training with the super varied, complete, entirety of human knowledge, is it?
If you somehow think it is though, I'd love to hear your reasoning.
There is no equivalency, only contributing factors. One cannot deny that our evolutionary history has contributed to our current capacity, probably in ways that are difficult to perceive unless you're an anthropologist.
Language is one mode of expression, and humans have many. This is another factor that makes humans so effective. To be honest, I would say that physical observation is far more powerful than all the bodies of text, because it is comprehensive and can respond to interaction. But that is merely my opinion.
No-one should be arguing that an LLM training corpus is the same as evolution. But information comes in many forms.
You're comparing the hyper specific evolution of 1 individual (an AI system) to the more general evolution of the entire human species (billions of individuals). It's as if you're forgetting how evolution actually works - natural selection - and forgetting that when you have hundreds of billions of individuals over thousands of years that even small insights gained from "running around in the jungle" can compound in ways that are hard to conceptualize.
I'm saying that LLM training is not equivalent to billions of years of evolution because LLMs aren't trained using evolutionary algorithms; there will always be fundamental differences. However, it seems reasonable to think that the effect of that "training" might be more or less around the same level.
Im so confused as to how you think you can cut an endless chain at the mouse.
Were mammals the first thing? No. Earth was a ball of ice for a billion years - all life at that point existed solely around thermal vents at the bottom of the oceans... that's inside of you, too.
Evolution doesn't forget - everything that all life has ever been "taught" (violently had programmed into us over incredible timelines) all that has ever been learned in the chain of DNA from the single cell to human beings - its ALL still there.
reply