ExpertsFMArtificial meat is harder than artificial intelligence — Lewis Bollard
Dwarkesh Podcast · August 7, 2025
5 Selected Lessons:
reasoning_pattern · 2 heuristic · 2 case · 1
Fixing one optimization problem spawns a cascade of new problems if the root cause isn't addressed
Source
Dwarkesh Podcast — Artificial meat is harder than artificial intelligence — Lewis Bollard (August 7, 2025) · at 13:56 ▸
When this applies
when a system is optimized along one axis and downstream symptoms are patched individually rather than addressing the root structural cause
The takeaway
When repeatedly patching symptoms of an optimization without touching root causes, expect an escalating chain of new problems, some worse than the original.
Why it works
Each patch addresses a symptom created by the prior patch or by the core structural decision, without removing the underlying driver (e.g., overcrowding, selective breeding), so new symptoms keep emerging.
One level up
True fixes require addressing root structural incentives, not just the latest visible symptom.
Where it stops holding
This stops working when a patch actually alters the root incentive structure, not just suppresses a visible symptom.
Detail
approach: Trace the causal chain from the root structural decision (e.g., crowding) through successive patches to see where each 'fix' just displaces the problem.
signals:
- each new fix narrowly targets the latest visible symptom
- the fix 'sometimes just makes it worse'
- root cause (overcrowding/selective breeding) remains untouched
rationale: Patching symptoms without changing the underlying incentive (efficiency at any welfare cost) perpetuates the cycle.
In their words
At each step, there is a new solution that that can't solve the fundamental underlying parts of the problem and sometimes just makes it sometimes just makes it worse.
Mislabeling destroys the market for genuinely differentiated ethical products
Source
Dwarkesh Podcast — Artificial meat is harder than artificial intelligence — Lewis Bollard (August 7, 2025) · at 56:06 ▸ · at 58:14 ▸
When this applies
when trying to build a premium differentiated product in a market where labeling/marketing allows commodity competitors to falsely signal the same differentiation
The takeaway
A premium differentiated product requires clear, trustworthy labeling to survive; in markets with rampant mislabeling, legitimate differentiation struggles to capture a price premium regardless of true quality difference.
Why it works
If consumers cannot reliably distinguish a genuinely improved product from a falsely-labeled commodity alternative, they won't pay the premium, starving the genuine innovator of the volume needed for viable economics.
One level up
Before entering a market with a genuinely differentiated/premium offering, check whether information asymmetry (mislabeling, vague claims) already undermines the market's ability to reward true differentiation — and advocate for labeling standards as a prerequisite.
Where it stops holding
This stops working when labeling is already clear/enforced (as Bollard notes happens in the egg sector with 'cage free'/'pasture raised' labels, where the premium market does grow).
Detail
approach: Check whether the product category has clear, legally meaningful labeling before assuming consumer willingness-to-pay will support a premium differentiated entrant.
signals:
- vague labels like 'all natural' mean nothing but are perceived by consumers as meaningful
- a $2 premium next to a falsely-equivalent cheap option kills adoption
- clear labels (cage-free, pasture-raised in eggs) correlate with a growing premium segment
rationale: If you're trying to sell your product as like pasture raised next to a product that says all natural and people think it means the same thing... you're not gonna get very far.
In their words
And unfortunately, they just went out of business. And I think the reason they went out of business is because there is such huge mislabeling across the industry that it's very hard to separate out what's actually better.
Neglectedness creates outsized philanthropic leverage
Source
Dwarkesh Podcast — Artificial meat is harder than artificial intelligence — Lewis Bollard (August 7, 2025) · at 27:44 ▸ · at 31:19 ▸
When this applies
when deciding where to deploy capital or advocacy effort among competing cause areas
The takeaway
Seek causes that are important, tractable, but neglected — the combination that produces an outsized ratio of impact per dollar — because saturated cause areas have already captured their cheap wins.
Why it works
When almost no capital flows to a cause because of low salience/popularity, the cheapest, highest-leverage interventions remain undone, so marginal dollars go much further than in saturated cause areas.
One level up
Capital efficiency in any cause area is inversely related to how much capital has already flooded it.
Where it stops holding
This stops working when the cause is neglected because it is NOT tractable (no effective interventions exist), in which case more money doesn't translate to proportional impact.
Detail
rule: Prioritize important+tractable+neglected causes for maximum marginal philanthropic leverage.
threshold: Farm animal welfare: <$300M/year globally, <$200M 'smart money'; climate advocacy alone is 50x bigger; cat/dog rescue in the US is 25x bigger.
rationale: This is a tiny amount of money for the purpose of social reform, and yet it has achieved a huge amount impacting hundreds of millions, billions of animals.
In their words
I think it's very unique. And I think the reason this philanthropic opportunity exists is because this area has been systematically neglected
Why corporate campaigns outperform legislative campaigns on low-salience consumer issues
Source
Dwarkesh Podcast — Artificial meat is harder than artificial intelligence — Lewis Bollard (August 7, 2025) · at 1:00:29 ▸
When this applies
when deciding whether to target a corporate decision-maker or a legislative body for a reform that has strong latent public support but low salience
The takeaway
For reforms with broad but low-salience public support, target corporate decision-makers directly rather than legislatures, because corporations face less entrenched capture and are more responsive to consumer-facing reputational pressure than politicians are to diffuse voter preference.
Why it works
Corporate decision-makers are less captured by a narrow, well-organized lobby than legislators (who face concentrated donor pressure on low-salience issues), and companies are more responsive to consumer sentiment than politicians are to voter sentiment on issues outside the voters' top-10 priority list.
One level up
Match your reform target (legislature vs. company vs. other institution) to where capture is weakest and responsiveness to your constituency is strongest.
Where it stops holding
This stops working when the target industry itself is the primary political actor being lobbied (i.e., the upstream commodity producer, which has no direct consumer-facing reputational exposure and is captured already).
Detail
rule: Target reputationally-exposed, consumer-facing corporate decision-makers over captured legislative committees for low-salience-but-popular reforms.
rationale: Politicians have decided that they need to be responsive on... the issues their voters care most about... on low salience issues like this, they can just ignore what their voters want... whereas corporations, if consumers are really outraged, then we need to act.
In their words
I also think companies have just proven more responsive to consumers than politicians are to their voters.
Commodity markup disappears once the higher standard becomes the baseline (Costco cage-free example)
Source
Dwarkesh Podcast — Artificial meat is harder than artificial intelligence — Lewis Bollard (August 7, 2025) · at 58:14 ▸
When this applies
when a 'premium' version of a commodity product is priced far above its true marginal cost because it hasn't yet become the market baseline
The takeaway
A quality/ethical improvement priced as a premium today will converge to the commodity baseline price once it becomes the mandatory or universal standard — advocacy should aim to make that happen (via commitment, not partial adoption), not just to create the premium option.
Why it works
Retailers can extract a markup from less price-sensitive consumers as long as the improved standard remains a minority/premium option; once a retailer (or a regulation) makes the improved standard the only option sold, competitive commodity pricing forces the markup back down to true marginal cost.
One level up
The fastest way to kill an exploitative markup on an improvement is to make the improvement the only option, forcing it back into commodity competition.
Where it stops holding
This stops working when the underlying cost of the improvement genuinely scales with volume in a way that keeps it structurally above the old baseline (then it never converges).
Detail
situation: Retailers pledged to go cage-free but many (Walmart, Kroger) stalled, citing high consumer prices for cage-free eggs ($1-2 more per dozen vs. 19 cents actual cost difference).
decision: Costco went 100% cage-free, eliminating the choice between caged/cage-free.
outcome: Costco now sells cage-free eggs at the same price as Walmart sells caged eggs, demonstrating the markup was exploitation of price-insensitive premium buyers, not a true cost pass-through.
lesson: Full commitment to the improved standard (not partial optionality) eliminates the price-discrimination markup and restores commodity pricing.
In their words
By contrast, Costco went a 100% Cagtree. They followed through on their promise. And what we see is they are now selling Cagtree eggs for the same price as Walmart sells its caged eggs.