What this covers

Ramez Naam, a technologist and science-fiction author who spent 13 years at Microsoft before founding a molecular design software company, joins Sean Carroll to argue that clean energy adoption is being driven primarily by economics rather than environmental concern or policy. The conversation centers on how solar and wind costs follow predictable learning curves—dropping roughly 30% with each doubling of manufacturing scale, a principle first observed in WWII airplane production—and have now reached the point where building new renewable capacity is cheaper than operating existing fossil fuel plants. Naam traces the mechanics of this decline: for solar, cost reductions come mostly from manufacturing improvements rather than fundamental technology breakthroughs, including higher labor productivity, automated machinery, thinner wafers, and cheaper inverters. The conversation moves across electricity generation, grid storage, and sectoral decarbonization, examining both the promise of technologies like flow batteries for long-duration storage and the hard problems where economics alone will not drive transition.

The harder terrain emerges when Naam addresses sectors without price competition: steel, cement, shipping, and cattle. Here he argues that policy must create markets through mandates—requiring that a percentage of steel or cement used in a region be carbon-neutral—to bootstrap learning by doing cycles in nascent clean technologies, following the model of Germany's subsidized solar deployment in the 2000s. Agriculture and land use, responsible for roughly a quarter of global emissions and driven heavily by cattle and deforestation, present a distinct challenge addressed through intensification and urbanization rather than direct decarbonization. The conversation touches on climate polarization in America, where clean energy enjoys near-universal public support despite climate change remaining deeply divisive, and explores how mobile money enables leapfrogging in Africa by making pay-as-you-go solar viable. Naam also reflects on narrative as persuasion and its power to shape belief beyond rational argument—a theme he explores through his science-fiction work.

Sharpest takeaway

Naam argues that the transition away from fossil fuels is being driven primarily by economics rather than policy or virtue—clean energy costs fall predictably along learning curves as deployment scales—giving grounds for optimism that we can decarbonize most of the economy, though hard sectors like steel, cement, and cattle still require policy-driven market creation.

  • Solar and wind costs follow predictable learning curves, dropping a fixed percentage with each doubling of deployment
  • Solar and wind are now reaching the point where building new clean energy is cheaper than running existing coal or gas plants
  • Sectors without price competition (steel, cement, cattle) need policy mandates to bootstrap markets and start the cost-reduction cycle

The argument · threads11 threads · 44 claims
0.85

Solar and wind costs are falling along learning curves, making clean energy cheaper than fossil fuels.

7 pointscentrality 5/5
  • Solar, wind, and battery costs fall along learning curves: every doubling of cumulative manufacturing scale has, on average, brought solar's price down by about 30%—an instance of Wright's Law, originally observed in WWII airplane production where each doubling cut cost by 16%.

    “they all follow what's called the learning curve or they have a learning rate or what's sometimes called a Wright's law. Wright was a World War II economist who studied the production of airplanes, and found that every doubling of manufacturing of the same model airplane brought down the price by a fixed percentage, by 16%.”

  • The cost of solar panels has fallen roughly 350-fold, from $77 per watt in 1977 to around 25 cents per watt today, and over just the past decade the all-in cost of solar electricity (land, labor, hardware, operation) dropped by a factor of ten.

    “In 1977, it was 77 bucks. And today, it's in the 20s of cents, 25 cents, maybe, let's say, like a 350 times cost reduction.”

  • Solar cost reductions come mostly from manufacturing process improvements rather than fundamental panel technology or R&D—higher labor productivity, bigger automated machines, thinner silicon wafers, less silver per watt, lower furnace temperatures, cheaper inverters—even though the solar industry reinvests only about 1% of revenue in R&D.

    “people criticize the solar industry 'cause the solar industry only reinvests 1% of revenue in R&D. But it's not the stuff on the R&D budget that's bringing the price down, it's actually factory efficiencies.”

  • Concentrating solar (solar thermal), which focuses sunlight to heat a working fluid like water or molten salt to drive a turbine—offering built-in thermal storage—has been overtaken by silicon-based solar photovoltaics that convert photons directly into electrons, with relatively few concentrating solar plants existing in the world.

    “that's an older form of solar called solar thermal or concentrating solar... they concentrate the sun's heat on a small area and heat up, usually water, which turns a turbine. Or more recently people are building projects like that with molten salt”

  • Naam installed solar on his Seattle home—one of the least sunny parts of America—and his utility bill dropped by more than the cost of the separate solar bill, even in summer months.

    “my bills are just going down, I have a separate bill for the solar. The utility bill has gone down by more than the bill for the solar, even in Seattle”

  • The growth in solar deployment will follow a sigmoid curve—accelerating once new clean energy beats existing fossil plants, then hitting headwinds from intermittency as solar and wind become a large share—but cost reduction may have no clear end point, since solar is only about 2% of world electricity and four more doublings (to 32%) could bring sunny regions to one cent per kilowatt-hour.

    “in the growth of solar, and how much solar we deploy, it is going to be sigmoid... But that's different than cost. It's not clear to me if cost really has an end point or not.”

  • Most solar deployment worldwide—about 70% in the US—is large centralized solar farms rather than rooftop, because centralized solar is generally higher value to the grid and captures economies of scale, though distributed rooftop solar gains value as distributed storage spreads.

    “In the US it's about 70% of our solar. And that probably does really have the highest value to the grid... centralized solar is just higher value, you can deploy it more cheaply and get the economies of scale than putting it on lots of individual roofs.”

0.76

Electrification plus grid storage can decarbonize most sectors if electricity becomes clean.

5 pointscentrality 5/5
  • The core decarbonization strategy, per David Roberts, is to electrify everything and make all electricity clean; emissions break down roughly into quarters—electricity (~25%), transport (~25%, with cars/trucks solvable but ships and planes hard at ~2% each), industry/building heat, and agriculture/forestry/land use (~25%).

    “He says, 'Look, the plan is this. Electrify everything and make the electricity all clean.'”

  • Grid storage will be roughly split between centralized (at solar/wind plants) and edge storage (homes, offices, malls, substations), because edge batteries provide extra benefits like outage backup and using local distribution lines that are full at peak (4-8 PM) but empty at midnight—filling batteries with cheap midnight wind and draining at peak—analogous to caching at the edge of a software network.

    “the electricity lines that run to your home in your neighborhood, they're probably pretty close to full at 4:00 PM to 8:00 PM but they're empty at midnight. So we can fill up the battery in your region at midnight with wind power ideally... and then drain it down at the peak of the day.”

  • A flow battery moves electrolyte through a membrane—one direction produces electricity, the other absorbs it—and its key advantage is that the size of the electrolyte storage tank can be separated from the power stack, so adding more hours of storage just means a bigger cheap plastic tank of cheap electrolyte (e.g. iron and salt), making long-duration storage very inexpensive.

    “A flow battery moves electrolyte through a membrane... in a true flow battery you can separate the size of the tank that contains electrolyte from the membrane in the power stack that controls how much power you can output.”

  • Lithium-ion battery cost dropped by a factor of nine between 2010 and now, and is expected to fall a further three- to five-fold, which would likely deliver cheap grid storage; alternative chemistries like flow batteries offer a step-function cost reduction better suited to the grid.

    “between 2010 and now, batteries, lithium-ion batteries... have dropped in cost by a factor of nine... Lithium-ion's going to drop in price another factor of three to five, and that'll actually probably be enough to get it to where we have really good cheap grid storage.”

  • Battery price has fallen far faster than energy density has improved—since 2010 price dropped by a factor of nine while energy density by weight rose only about 60-70%—because, as Elon Musk puts it, one engineer improving the battery factory is worth ten engineers improving the battery technology.

    “since 2010, the price has come down by a factor of nine, the energy density by weight has gone up by maybe 60%-70%”

0.76

Clean energy adoption is driven by consumer preference and business economics, not climate ideology.

5 pointscentrality 5/5
  • Climate change is hugely divisive in America (uniquely so among countries) and that polarization has not changed, yet clean energy is almost universally loved—super-majorities everywhere rank solar as their first choice and wind second, with natural gas far behind and coal last—which is what actually drives clean energy progress rather than climate concern.

    “climate change is hugely divisive, and that hasn't really changed... But super majorities everywhere say that solar is their first choice for energy source for their community, wind is their next, natural gas is like 20 points down from that, and coal is down at the bottom.”

  • NIPSCO, a Northern Indiana utility that is 65% coal-powered in a red state that voted for Trump by 19 points, announced a five-year plan concluding it would save customers $4 billion to shut down all its coal and replace it with solar, wind, batteries, and flexible demand—demonstrating the disruptive crossover is happening now.

    “this utility in Northern Indiana called NIPSCO announced their five year road map... they are currently 65% coal-powered, they're in a red state, went for Donald Trump by 19 points. And they said that it would save their customers $4 billion for them to shut down all that coal and replace it with solar, wind, batteries and flexible demand.”

  • In some places it is now cheaper to build entirely new solar, wind, and storage than to keep running existing coal—and in some cases existing natural gas—power plants, marking a disruptive third phase of clean energy.

    “it's cheaper to build new solar and wind and storage in some places than it is to keep existing coal power plants. And in some cases, existing natural gas power plants running.”

  • Within two or three years a new electric car with 200+ mile range will cost the same or less than a comparable internal combustion car while costing about a quarter as much to run (maintenance and energy per mile), making EV adoption a no-brainer; range anxiety is now more psychological than real given vehicles like the 325-mile Tesla Model 3.

    “in a couple of years, maybe two years from now, maybe three, a new electric car... with a 200 mile range or more, will be the same price or cheaper than an internal combustion car that's very similar, and it'll cost a quarter as much to keep running.”

  • Any increase in electricity demand that can be located anywhere on Earth—such as Bitcoin mining—is actually good for the planet, because it scales solar and wind (the least-cost sources) faster, driving them further down the learning curve, even though Bitcoin miners in China are constrained by money-control evasion rather than choosing the sunniest sites like Chile.

    “I think that an increase in electricity demand, if you can put your electricity demand application anywhere on Earth, is just good for the planet. 'Cause it allows us to scale solar and wind faster, 'cause they're least cost, and that drives them down in cost faster.”

0.63

Agriculture, livestock, and deforestation are major emission sources requiring targeted technological solutions.

5 pointscentrality 4/5
  • Agriculture, forestry, and land use is roughly a quarter of emissions, dominated by mammal livestock (overwhelmingly cattle) and deforestation; cow burps of methane alone are about 4% of global emissions—bigger than all aviation and shipping combined—and cattle have historically been the number one driver of deforestation.

    “Cow burps of methane... Are about 4% of global emissions. So they're as big as planes and ships put together”

  • Deforestation harms the climate twice: it removes a carbon sink and immediately releases the carbon already stored in the trees and plants, especially when poor farmers in Indonesia/Malaysia use fire (cheaper than sawing) to clear land—fires that on some days have emitted more carbon than the entire economy of Japan.

    “Takes away a sink, and more than that, the carbon that's already in that forest, that's in the trees and plants, gets emitted into the air... So you produce an immediate emission that's quite large.”

  • Feed additives—most famously a red seaweed that in small experiments massively cut cattle methane—may reduce cow burps, and because methane is high-energy, the energy retained could turn into more cow mass or milk; if validated (the skeptical science is not yet done), this offers a path that helps farmers and the planet, with microbiome engineering as a fallback.

    “there's a red seaweed that in very small experiments, when cows ate it, massively cut their methane production... if you feed a cow this stuff, not only does it produce less methane, but that energy, methane is a very high-energy molecule, that energy gets turned into more productivity.”

  • A third of the planet's land is agriculture, a third unusable, a third forest, with human habitations only 2-3%; shifting people from agrarian to urban economies (densification) plus intensifying agriculture—Bangladeshi yields are a quarter of US yields, so giving farmers tractors, fertilizer, pesticides, and GMOs raises productivity—lets us pass laws to spare and protect forests while still feeding people and lifting them out of poverty.

    “A third of the land area of the planet is agriculture, a third is unusable, mountains, deserts, so on, and a third is forests, more or less... cities are like... Human habitations are like 2% or 3%.”

  • Plant-based and stem-cell meat substitutes (Impossible Burger, cultured meat) have a real chance to displace processed meat where original texture is unimportant—so 'Impossible pepperoni' arrives before 'Impossible T-bone steak'—but only beef matters much for climate, so this is a partial, not complete, solution.

    “when you look at processed meat where you don't have the original texture, I think they have a chance to go into that.”

0.58

Industrial sectors without price competition need policy mandates to create markets for clean alternatives.

4 pointscentrality 4/5
  • For sectors without price competition—steel, cement, planes, ships—we must replicate what policy did for solar and wind: deploy mandates (e.g. requiring X% of cement or steel used in a state to be carbon-neutral) to create a market that lets startups raise funding, scale, and ride down the learning curve until clean methods undercut the old ones, just as Germany's subsidized solar deployment in the 2000s started the global cost decline.

    “what we need to do is do that with the sectors that we don't yet have price competition for, is we have to say, 'We're going to have a mandate. X percent of the cement... has to be carbon neutral.'... And that will create a market, that will allow the startups that have technology that can do that to raise funding”

  • Steelmaking produces carbon emissions for two distinct reasons: needing a cheap heat source (currently burning gas or coal, which clean electricity must reach about 1.5 cents/kWh to beat) and process emissions, where coal acts as a reducing agent supplying carbon atoms that bond with oxygen in iron ore to form CO2; substituting hydrogen as the reducing agent yields water instead, though at higher cost.

    “in steel, iron ore contains a whole lot of oxygen and you can't have oxygen in your steel, so you have to have a reducing agent that takes the oxygen out. And right now, you burn coal and it supplies both the heat and the carbon atoms that combine with the oxygen and they turn into CO2”

  • Although solar is cheaper, wind is valuable because it is counter-cyclical to solar—wind is statistically faster at night and peaks in spring/winter when solar is lowest, while solar peaks in summer when wind is at its trough—so a continent-sized grid combining both at ideal locations can supply 70-80% of electricity without storage.

    “the nice thing about wind is it's counter-cyclical to solar, roughly... Wind statistically is faster at night than it is during the daytimes.”

  • Seemingly incremental policy can do a lot: California's renewable portfolio standard started modest (~20-25%) and is now 100%, and requiring even 20% of cement and steel entering a state to be clean—which sounds achievable—would create a vastly larger market for carbon-neutral materials than exists today and help those industries scale.

    “California... I think it was 25% or 20% by a certain year, and now it's 100%. But getting up to 20%, that doesn't sound crazy. If we said 20% of cement and steel coming into California... has to be clean, that doesn't sound unachievable. But that would be a gigantically larger market”

0.58

Global connectivity and smartphone ubiquity enable leapfrogging and economic development in poor regions.

4 pointscentrality 4/5
  • Rather than a technological singularity, Naam believes in the 'synchularity'—the moment when everyone on Earth is connected—which is arriving now because a cell phone (and cell tower) is cheaper than a toilet (and plumbing/waste system), enabling poor populations in India and Sub-Saharan Africa to leapfrog directly to smartphones and internet before modern sanitation.

    “I don't believe in the singularity, but I do believe in the synchularity when everyone is connected... a cell phone is cheaper than a toilet. And a cell phone and a cell tower is cheaper than a toilet and plumbing and a waste processing system.”

  • Mobile money is the key to getting solar onto homes in Africa: because it is too expensive to send agents around to collect cash, software-controlled pay-as-you-go solar panels (which shut off if you haven't paid) become viable, and where mobile money exists solar is cheaper, cleaner, and safer than kerosene, creating better educational outcomes and home income—a domino effect toward economic growth.

    “having mobile money of some sort allows you... Is the key to getting solar on the roof of your home in Africa. Because it's too expensive to send somebody around every place to collect cash. But if you have mobile money, the solar panels are literally software controlled.”

  • For most of the planet the smartphone is the first screen (not the third as in the US), making it the platform for education, transparency (photographing corrupt officials), market price information, and financial inclusion; cell phone penetration across Africa is in the high 80s-90s for adults with smartphone penetration racing toward ubiquity within a few years, which could be one of the greatest poverty-reducing developments on Earth.

    “It's the first phone for most of the planet. It's the first screen for most of the planet. So that becomes the platform for education, that becomes the platform for transparency... knowing market prices... financial inclusion, all of it.”

  • Effective development aid bootstraps local industries rather than handing out goods: handing out food in non-emergencies suppresses local food prices and drives farmers out of business, so it is better to provide seed capital, advice, better seeds, and agricultural techniques, plus the institutional preconditions—free markets, competition, access to capital, paid debts, low corruption, and adequately paid civil servants.

    “Handing out food in an emergency, for instance, is terrible, 'cause it suppresses the local prices of food, the farmers can't make any money, it drives the farmers out of business... you want to provide seed capital and advice for people to be able to build businesses on the ground”

0.54

Narrative and empathy shape human behavior more effectively than rational argument or data.

3 pointscentrality 4/5
  • Fiction and narrative are a 'super power' for changing opinions because human brains did not evolve to be rational or data-weighing; a narrative slides past rationality and penetrates a deeper level of the reader or listener, so it must be used responsibly because it can make people believe things deliberative argument cannot.

    “I do think fiction narrative is a super power, actually, because human brains did not evolve to be irrational or data weighing... if you can tell a narrative, be very responsible with it, because it slides past rationality and penetrates a deeper level of the reader or listener and they can start to believe things.”

  • Pervasive video capture and automatic real-time speech translation (in the speaker's own voice) will tap into and increase human empathy—seeing police-brutality videos rather than reading a statistic puts you there, and being able to consume foreign media or talk to an Iranian pen pal instantly, without the alienating effect of subtitles or dubbing, could expand who we count as part of our in-group.

    “These videos tap into human empathy. If you read a stat about police brutality, it's one thing. If you watch some of the videos, you're really there.”

  • Rational is not the default state of a human—tribal is—and empathy dissolves tribalism or at least expands one's sense of who belongs to the in-group, which is precisely what is needed to address global problems.

    “rational is not the default state of a human, tribal is, and empathy dissolves tribalism or at least expands your sense of who the in group is and that's what we have to do with global problems”

0.50

Human history shows measurable progress across health, poverty, and violence despite media bias toward negative news.

2 pointscentrality 4/5
  • Human history shows measurable improvement: bad things like hunger, child mortality, disease, and warfare are declining while good things like life expectancy and education are rising, providing an empirical basis for optimism.

    “I'm in the Max Roser, Steven Pinker camp of look at the numbers, all the bad things, hunger, child mortality, disease, warfare going down and good things, life expectancy, education, etcetera, etcetera going up.”

  • Because news media's business model rewards outrage and the most negative stories get the most revenue, readers should actively seek the larger context, rates, frequencies, and fractions behind any bad event (e.g. the rate of a salmonella outbreak versus elsewhere)—and people should seek out and share the ways the world is getting better, since that motivates rather than breeds complacency.

    “the news exists to get your clicks and that's their business model, they know that outrage works... the story that gets the most revenue for news media is the most negative story. And so... go find the ways the world is getting better and tell people about it”

0.36

Tools that amplify collaborative capability and reduce communication friction accelerate collective discovery.

2 pointscentrality 3/5
  • More intuitive tools for sharing ideas, knowledge, and designs amplify collective intelligence: modern discovery is done by giant collaborating teams, not lone geniuses, so reducing friction and increasing bandwidth between people (exemplified by Slack reducing email and preserving searchable team history) lets such groups work more effectively.

    “it amplifies our collective intelligence... it's giant teams of people working with giant instruments that were designed by giant teams of people that do the work. And so that's collective intelligence... if we have more intuitive ways to share ideas, knowledge, questions, designs... that reduces the friction and increases the bandwidth between us”

  • Direct brain-computer interfaces (Neuralink, Kernel, OpenWater) are being actively pursued but will arrive slowly; the intermediate human-augmentation step, perhaps 10-20 years away, is a combination of augmented reality glasses plus speech and natural language—essentially an always-present, far better Alexa that can see your surroundings and show you relevant data without spamming you.

    “People are really working on next generation of brain-computer interfaces, but it will be slow. I think the layer between... smartphones and directly in your brain is probably some combination of augmented reality and speech and natural language.”

0.27

Connected technologies create new risks of abuse that society must address through norms and regulation.

3 pointscentrality 2/5
  • Fundamentally, anything connected can be hacked, so brain-computer interfaces and other connected augmentation technologies open new avenues for abuse.

    “Anything that you can hack you, anything connected can be hacked, fundamentally.”

  • Privacy is not dead (contra David Brin); rather society will negotiate a new, messy social contract around pervasive recording devices—creating places where such devices cannot operate or record, with hacking devices to ignore those signals becoming a serious social faux pas—just as strong social conventions, not mere possibility, keep people from driving cars into crowds.

    “I don't think privacy is dead. I think we'll negotiate a new social contract, and I think it'll be messy. I think you'll have places where devices like that can't operate, or can't record at any rate.”

  • Contrary to the view that younger generations have abandoned privacy, young people understand privacy deeply—perhaps more than adults—wanting to control who they communicate specific things to (e.g. anything about their relationship to sex), even if they share more openly on certain topics.

    “I don't think young people have gotten out of the idea of privacy. Young people very much understand privacy and they're maybe more into it than adults. Ask a young person about anything about their relationship to sex and you'll find out how they feel about privacy.”

0.27

Climate warming to two to three degrees is serious but not apocalyptic; adaptation alongside mitigation matters.

3 pointscentrality 2/5
  • Business-as-usual climate projections used to point to 6 degrees Celsius of warming, but that curve has been bent and we are now tracking toward two-and-a-half to three degrees—better than before but still short of staying below 2 degrees, so more action is needed.

    “Now, it used to be that business as usual was 6 degrees. We've bent that curve, and now I think we're tracking for two-and-a-half to 3 degrees, but we should do more.”

  • Even though models say feedback loops like permafrost methane release are unlikely to kick in at 3 degrees and remain small contributors so far, there are tail risks of positive feedback loops, and it is unsettling to see non-linear processes heading in the wrong direction, especially when somewhat unpredictable.

    “there are tail risks, there are positive feedback loops that most models say are unlikely to kick in at 3 degrees C. But, they could.”

  • Two degrees of warming is a large amount—warmer than at any point during human existence as a species—but it is not a threshold where the world ends; 1.9 degrees is not fine and 2.1 degrees is not Armageddon, so adaptability matters alongside mitigation.

    “it would be warmer than the planet has ever been since humanity existed as a species. And if you think the middle of the last ice age was only about 4 degrees Celsius colder than it is now”