The skeptical case: a top-down benchmark against yeast

Why it has its own page

The workshop group was largely drawn from people already working in or close to the field. In the final session one participant described the group as "kind of the believers", and researchers who have published pessimistic analyses were not well represented live (scope note). We have since invited forecasters and modelers without ties to cultured meat, including people who had previously given skeptical forecasts. This argument came from that round.

Until now it was split across a folded section of the crux map, two folded paragraphs in the synthesis, and a long passage in the beliefs analysis. Collecting it here makes it easier to find and easier to answer.

The benchmark: yeast fermentation

Most substantial responses here reason bottom-up: project media cost, cell density, capex and scale, then multiply through. This respondent argues that approach is the problem. It combines many uncertain assumptions with unknown interactions, so it is highly sensitive to optimistic inputs. They anchor instead on a technology that already exists at industrial scale.

Yeast doubles in roughly 1.5 hours, grows on very cheap media, tolerates harsh industrial conditions, and scales on mature low-cost equipment. Animal cells do none of these things. The respondent estimates the total production-cost disadvantage at about three orders of magnitude (two if one is very optimistic): roughly two from volumetric productivity alone (lower density combined with roughly 15 times slower doubling, which sets reactor volume and hence capex and opex per kg), plus about one from growth factors and recombinant proteins that microbial fermentation never has to buy.

The empirical anchor. Industrial yeast fermentation is one of the most optimized bioprocesses there is, and commercial yeast-derived single-cell protein still prices in roughly the same order of magnitude as animal meat. The inference: if the most efficient biological production system available has not become dramatically cheaper than meat, it is hard to believe cultured animal cells, starting from a much less favourable position, reach commodity food cost. Asked whether a cultured-versus-fermentation gap would settle the funders' question if fermentation itself sat above meat cost, the respondent answered directly: they read fermentation single-cell protein as already near meat parity, so cultured meat sits well above animal meat rather than near it.

The mechanisms

The argument rests on specific causal claims, not general pessimism. Each can be argued with separately.

More detail, and a form to respond to each mechanism: mechanisms behind the top-down critique.

A point that applies to every estimate here

The focal question (CM_01) asks about undifferentiated cell biomass at harvest. The respondent notes that a real meat product also needs a differentiation phase, in which cells stop dividing but keep consuming nutrients, plus scaffolding and work on taste and texture. None of that is inside the CM_01 boundary, and the proliferative numbers from CHO-cell biopharma that several optimistic estimates lean on ignore it. If they are right, every figure on the site, optimistic and pessimistic alike, understates the cost of finished cultured meat by an unknown but non-trivial factor.

They also point to overflow metabolism (glucose shunted to lactate) as a reason a cheap medium per litre can still give an expensive kilogram, and argue that feed conversion should be modelled explicitly rather than left implicit. A second independent respondent, reasoning from nutrient consumption, reached a similar conclusion: the cost-driving quantity for media is yield, media consumed per kg of output, while density matters mainly for reactor volume and capital. We are considering rewording one of the questions for optimists because of this (note on the crux map).

Why a gap could persist

Separately from the biology, the respondent argues that a large sustained gap is partly self-perpetuating. While cultured meat sits far from parity there is little private incentive to fund the deep, coordinated, largely non-appropriable R&D needed to reach even its biological floor, so that floor may never be reached. The field compounds this by fragmenting: funding is spread across companies that each develop and characterize their own cell line. And because cost is a product of many drivers, isolated wins such as cheaper growth factors do not add up to a low headline cost unless media, productivity, senescence and capital improve together.

For funders this is an argument about what coordinated funding would have to look like, not only about what the technology can reach.

What would change their view: feed-grade growth factors at scale, several productivity breakthroughs, or economical large-scale perfusion. Where they would put the money instead: plant-based foods and microbial fermentation, which they see as offering higher expected returns per research dollar. Not because cultured meat is uniquely hard, but because those platforms are already biologically efficient and industrially validated while remaining, in their view, scientifically immature and open to innovation.

Three questions for optimists

The respondent put these to the optimists. They are answerable, and answers would separate the positions instead of restating them.

Where it sits among the estimates

The four independent 2036 point estimates are $6, $40, $50 and $500/kg; the ten standardized workshop estimates have a geometric mean of $17.7/kg. The $500/kg estimate is roughly five times the next-highest. Its effect on the pooled figures, and every individual estimate with its interval, are in the beliefs analysis.

With so few independent responses the comparison is descriptive. The respondent's number matters less than the method: it is the only response that tries to bound cost from the top down.

Published work does not settle it

The yeast comparison and the orders-of-magnitude breakdown are arguments to test, not verified results. Relevant published analyses range from the more optimistic continuous-production analysis of Pasitka et al. (2024) to the more skeptical scale-up analysis by Humbird (2021), with the Goodwin, Aimutis and Shirwaiker (2024) scoping review and CE Delft's scenario analysis in between. See the full source list. These frame the disputed assumptions; citing them does not validate any respondent's forecast.

The other side, and what we are doing with it

If you work on this, the most useful thing is a short answer to one of the questions above, or a skeptical argument we have not yet heard. Both go through the crux response form; you can respond publicly, anonymously, or privately. You can also annotate this page with Hypothes.is.