[dredmorbius]

Thanks for that.

I'd argue that the numbers suggest that venture spend on info tech / AI / social media is at least of the same order of magnitude as new drug discovery, if not more.

It would be interesting to compare blockbuster drugs of the past 5--10 years or so to see what types of investment have been made there. mRNA / Covid19 vaccines, and GLP1s.

A Deloitte report gives oncology, infectious disease, and chronic conditions (particularly obesity and diabetes) as key areas.

It also gives some relevant numbers re investment and revenue:

The increase in IRR is driven by promising late-stage pipeline candidates and impressive trial outcomes (the average forecast peak sale has increased to US$510 million). However, high R&D costs, which reached an average of US$2.23 billion per asset in 2024, present a continuing challenge.

<https://www.deloitte.com/ch/en/Industries/life-sciences-heal...>

(And if you're confused by exactly what that 'graph is actually communicating, I am too: "$2.23 billion per asset" is an odd and unclear phrasing.

From StartUs Insights, "Drug Discovery Market Report 2025":

The global drug discovery industry was valued at approximately USD 106.70 billion in 2025.

(Emphasis added.)

And...

The drug discovery sector includes 5370 companies with 990+ startups. The leading country hubs are the USA, the UK, India, China, and Japan, and major city hubs include San Diego, Cambridge, London, New York City, and San Francisco. ...

Over 9550 funding rounds closed with an average of USD 34 million per round. More than 7600 investors invested in over 2220 companies.

Doing the maths on that last: about $325 billion in funding.

<https://www.startus-insights.com/innovators-guide/drug-disco...>

Also relevant is projected growth of industries.

AI, which leads infotech venture spend presently, is valued at $515 billion in 2026, and is projected to grow at 30.6% CAGR, reaching $1 trillion in 20209 and $3.5 trillion by 2033.

<https://resourcera.com/data/artificial-intelligence/ai-marke...>

(Whether or not you, or I, believe these projections is less material than what the markets and investors believe, and they seem to be acting on the assumption that the market and asset values will grow.)

Pharma by contrast, is estimated at $1.7 billion 2025, growing at 6.08%, and reaching $2.8 trillion by 2033.

<https://www.grandviewresearch.com/industry-analysis/pharmace...>

Which is to say that whilst projected AI and pharma markets in five years are comparable, growth in AI is seen as 5x the rate of pharma. And investment follows growth, not total value.

Pharma spend is largely capped by public and private insurance spend, which tends to be mature, already high relative to total economic activity, is constrained on multiple fronts, and divides results and rewards amongst multiple entities (both drugs/therapies and companies). Biggest growth is projected in Asia-Pacific.

Projections for AI spend are ... somewhat unhindered by empirical data or practical realities, and at least as told at investor storytime, promise both the heavens and stars and winner-take-all dynamics, such that money is chasing the possibility of landing on the top of the heap.

Again: quite probable that there's a lot more money available for AI moonshots than new drug discovery, though perhaps grounded more in anticipation and psychology than reality or net social benefit.

[Retric]

> at least of the same order of magnitude as new drug discovery

That’s a valid way of looking at things, but diminishing returns tells a different story.

People have been looking at treating the same disease with 100’s of billions / year across decades. What completely new ideas haven’t been explored with the 10’s of trillions of dollars we already spent? Vs deep learning or perhaps more narrowly LLM style AI where the total investment over the last 50 years is probably ~1% of what’s been spent on drugs.

So sure in 2026 things look heavily weighted to “junk” tech, but overall it’s probably at least a 2 orders of magnitude difference from a slightly different but still reasonable take. Similarly looking forward people are only going to toss 100’s of billions at it in 20 years if it’s actually had a massive impact, without that it’s a short term blip.

[dredmorbius]

Another valid line of argument.

Increases in healthcare spending have had remarkably little outcomes improvement. And that's been the case for a long, long, time, 50--60 years by some arguments. Robert Gordon made a strong case for this in his book The Rise and Fall of American Growth (2016), citing the work of healthcare economists, notably Victor Fuchs.

But ...

... "You can see the computer age everywhere but in the productivity statistics", as Robert Solow stated. Computer capabilities have increased roughly 33 millions-fold since the 1950s, but productivity hasn't followed that same curve. Even looking at information technology spend vs. growth shows limited gains (computer capacity has become vastly cheaper over the same period). It seems likely that AI will follow a similar trend.

The one technological domain in which there was a near-linear relationship between increased input and output was mechanisation. Going back to Gordon, two of the most interesting plots in the book were of horsepower and kW per worker vs. output. The relationship wasn't quite 1:1, but it was strongly linear, in that an n-fold increase in input power provided a k*n increase in output (with k < 1, but modestly so). That is ... rare ... in technological progress. Far more often we see dramatically falling returns to scale.

Back to medicine, two of the most remarkable areas of improvement in the past decade have been mRNA vaccines and GLP1s. Granted the latter are an older technology (by several decades), but improvements permitting once-per-week dosing and oral dosing, rather than daily injections, have dramatically increased availability of GLP1s, and their benefits across a huge range of conditions are remarkable. Probably not so much as general hygiene, sanitation, and safety improvements from say 1850 -- 1925, which saw the most remarkable reductions in human mortality. But respectable all the same. Given the growing threat of novel infectious disease, mRNA vaccine development within days of an outbreak's announcement is absolutely stunning. I'll tip my hat to health research there, and note that fundamental research in both cases languished until quite recently.