Exponential Growth: Why AI, Solar & Batteries Will Keep Getting Cheaper

Exponential View & Cleaning Up | Azeem Azhar & Michael Liebreich
Release Date: November 28, 2024

Episode Overview

In this special crossover episode, Azeem Azhar (Exponential View) joins Michael Liebreich (Cleaning Up) for a robust, incisive discussion about the transformative, exponential trends reshaping the global energy and technology landscape. Together, they dissect how AI, solar, and batteries are defying historical cost constraints, the mechanisms fueling exponential growth, and the frictions—technical, regulatory, and economic—that could throttle or redirect this trajectory.

With deep experience as analysts, investors, and thought leaders, both hosts navigate optimism, skepticism, and the tangible lessons of history, scrutinizing whether the rapid decline in the cost of key exponential technologies will continue, and how these shifts interact with business, politics, and society at large.

Key Discussion Points & Insights

1. The Drumbeat of Exponential Change

[03:23–08:59]

• Azeem Azhar's Background & Epiphany:

  • Grew up in the microprocessor era, internalizing Moore's Law from an early age.
  • Noted exponential adoption curves in internet and computing, later recognizing similar patterns in solar, batteries, and genome sequencing.
  • “What I discovered…was that it wasn’t just silicon chips that have this characteristic of exponentiality…There were technologies like solar panels, technologies like lithium ion batteries that were getting cheaper and cheaper and cheaper.” – Azeem [07:58]

• Learning Curves and Wright's Law:

  • Both discuss the foundational concept that as cumulative production rises, costs fall—sometimes dramatically (Wright’s Law, 1936).

• Historical Evidence:

  • Michael recounts: “Anybody who has bet against them being exponential has lost. That’s the important point.” [10:01]

2. Solar & Batteries: Real-World Exponential Decline

[08:59–14:47]

• Price Collapses:

  • Computing: 1 gigaflop in 1961 = $190 billion; in 2023, one and a quarter cents.
  • Solar: $130/watt (1975) to $0.31/watt (2023).
  • Batteries: $7,500/kWh (1991) to <$50 (today).

• Solar’s Scaling Analogy:

  • Solar installation timeline: annual → monthly → weekly → daily gigawatts.
  • “Where we are today, you’ve now got the Chinese solar industry, which can produce enough Solar to produce 2 terawatts of power…10% of all the electricity that we use today could be replaced with one year’s output of Chinese Solar manufacturing.” – Michael [12:56]

• Modularity & Democratization:

  • Modular nature of new tech enables small players—from homeowners to SMEs—to enter energy markets, expanding the addressable market and accelerating deployment.
  • “What’s happened with solar PV in particular is it’s transformed who can access the energy system.” – Azeem [15:14]

3. Exponentiality’s Limits and Constraints

[20:28–32:58]

• Pushback 1: Not Everything Is Exponential

  • Michael flags hydrogen as an example—cheapening electrolyzers do not make overall green hydrogen cheap, due to system costs (storage, distribution).
  • “The electrolyzer stacks for green hydrogen get cheap. That they could go to zero, it doesn’t matter. But they are only…11% of the cost of green hydrogen today.” – Michael [22:34]

• Pushback 2: Whole-System Costs & Entrenched Regimes

  • System costs, regulatory capture, and incumbent interests can slow or block transformation.
  • “None of this stuff can continue to happen…regulators answer to political masters…and they are paid for, bought and trussed up like Christmas turkeys by the oil and gas industry.” – Michael [27:07]
  • Azeem points to the momentum of markets: “Ultimately, you have the law of physics, then you have laws of economics, and then the third line is just the law of reality. And if the electricity market is not functioning, governed as a fossil fuel market…lobbies will not be able to stand in the way.” [29:07]

• Pushback 3: Concentration and Inequality

  • Even in democratized technologies (e.g., solar), value and power can concentrate—e.g., dominant Chinese manufacturers, tech monopolies.
  • “At the end of this…we’re just going to have…colossal Chinese companies dominating everything. We are not going to be any freer at the end of this than we were before.” – Michael [34:18]
  • “The Chinese companies making solar are much more like the people who made RAM chips in the computer industry…They compete on price…It’s a super commodity.” – Azeem [34:55]

• Pushback 4: Stability & Systemic Risk

  • Distributed, modular systems create complexities—particularly for critical, continuous-use customers (aluminum smelters, hospitals).
  • “How do you keep the lights on, when everybody’s got their very cheap system, which is fine, 48 weeks a year…but on the four weeks a year that it’s not fine, they turn around and say, ‘What’s the government doing about the fact my lights don’t go on anymore?’” – Michael [36:41]
  • “We may have to keep a bunch of gas plants, you know, idle as insurance…” – Azeem [38:46]

4. Data Centers, AI, and Energy Demand

[42:18–57:57]

• The “AI Power Crunch”:

  • Explosive growth in AI and generative applications driving demand for gigawatt-scale data centers, raising questions about electricity supply and energy infrastructure.
  • “The AI companies have put in orders for 500 megawatt, 700 megawatt data centers and they’re talking about gigawatt and 5 gigawatt ones in a few years time…” – Azeem [46:39]

• Reality Check on Supply:

  • AI use so far: incremental data center build is “about an additional 20%” to traditional [43:37], but huge growth forecasted.
  • Investment: Data centers ($140B/yr → $250B by ‘27), corresponding $200B required for incremental power system investment [54:16–54:52]—but compared to $2T/yr global renewables, it’s significant, but not overwhelming.

• Limits of Exponential Growth in Practice:

  • Building new nuclear, grid connection, or even mega data centers presents practical bottlenecks—permitting, land, build times.
  • “Some time horizons don’t add up…that timeline will not be as close as three or four years. It will get pushed out a couple of years. That’s just the reality.” – Azeem [53:16]
  • “I think that the things that they think they’re going to be able to deliver by 2030 are probably 2035.” – Michael [56:15]

• Sustainability of Demand & Markets:

  • Discussion of potential correction as overbuilding meets real-world limits—“Aren’t we going to see a correction at some point along the way?” – Michael [59:10]
  • “We are bound to see a correction. The question is what type will it be?” – Azeem [59:30]
  • AI demand is real, not purely speculative, but cycles and shakeouts are certain.

5. The Underlying Question: Infinite Growth, Finite Planet

[63:07–68:57]

• Is Infinite Growth Possible?

  • Michael invokes Herman Daly: “There’s no such thing as infinite growth on a finite planet.” [63:17]
  • Azeem’s perspective: “The climate crisis will get addressed in large part by these technologies…But there’s a bigger question about this idea of infinite growth on a finite planet…There’s a lot that we don’t know…maybe we don’t have to get that far.” [64:07]

• Choices and Values

  • “We get to set some of these values and we get to articulate them…I think there’s been three points where you’ve been more optimistic than me…you think that we will generally use these technologies for the good rather than for the bad. And I think where I am…is that that’s a choice that we get to make.” – Michael [69:05]

Notable Quotes & Memorable Moments

• On Exponential Experience Curves:
  “If you look at the total amount of computing capability on the planet…you see a 62% compounded annual growth rate…for over 51 years. So yes, at some point that exponential may stop. But 51 years…is enough to transform economies and transform the way we live.”
  – Azeem [10:21]

• On Solar’s Market Compounding:
  “Today, homeowners, businesses…can get into this market and it really expands the market much, much faster than ones that have been constrained by…an older technology.”
  – Azeem [15:14]

• The Law of Physics vs. Economics vs. Reality:
  “Ultimately you have the law of physics, then you have laws of economics, and then…just the law of reality.”
  – Azeem [29:07]

• Monopoly Shifts:
  “We are not going to be any freer at the end of this than we were before.”
  – Michael [34:18]
  “Solar PV is much more like DRAM than it is like chips…They’ll be the biggest producer of a pile-it-high, sell-it-cheap product.”
  – Azeem [34:55]

• On Overbuilding and System Stability:
  “There is not an asset class that we manage for 100% utilization…when it comes to renewables people say…it doesn’t work, it’s rubbish.”
  – Michael [38:46]

• On AI’s Energy Demand:
  “Microsoft and Google have both talked about spending more than $100 billion to develop the state-of-the-art models…the cost of not getting this right is existential.”
  – Azeem [47:37]

• On Market Cycles:
  “We are bound to see a correction…the question is what type will it be?”
  – Azeem [59:30]
  “The Nasdaq took 20 years…to recover from the Internet boom.”
  – Michael [59:56]

• On Infinite Growth on a Finite Planet:
  “Maybe we don’t have to get that far. But I like to play the thought experiment…if you were in 1750 and you were presented with that question, you probably wouldn’t believe we could do what we did by 1950.”
  – Azeem [64:07]

• On Societal Choices:
  “That’s a choice…It will not be made by Elon Musk or by the CEOs of the tech titans. It will be made by society as a whole…we have to be prepared to fight for [it].”
  – Michael [69:24]

Segment Timestamps

• 00:00–03:23 – Introduction & Host/Guest Backgrounds
• 03:23–08:59 – Exponential Change: Technology Learning Curves
• 08:59–14:47 – Historic Price Collapses: Solar, Batteries, Computing
• 14:47–20:28 – Modularity and Democratization of Technology
• 20:28–32:58 – Limits to Exponential Growth: Hydrogen, System Costs, Regulation
• 32:58–38:46 – Systemic Frictions: Market Concentration, Stability Risks
• 38:46–42:18 – System Utilization, High-Demand Users & Subsidy
• 42:18–57:57 – Data Centers, AI, and Energy Infrastructure Bottlenecks
• 57:57–63:07 – AI’s Economic Model, Market Corrections
• 63:07–68:57 – Infinite Growth vs. Finite Planet, Societal Values
• 68:57–End – Final Reflections and Closing

Episode Tone

A balanced, probing, candid, and occasionally humorous back-and-forth, blending Azeem’s optimistic but data-driven enthusiasm with Michael’s engineer-pragmatist skepticism—a kind of debate between vision and caution, future shock and present constraint. Both share a sense that the ultimate direction of technological transformation will be determined not just by physics or economics, but by societal values, choices, and the capacity to adapt regulatory frameworks to fast-moving markets.

Summary Prepared For: Listeners Seeking Insightful, Actionable, and Nuanced Perspectives on the Future of Energy, Technology, and Society.