tl;dr: the eventual collapse of "safe and easy" SaaS investment will force VC to take actual risks for the first time in a while. i'm quite excited to see the results.
this is an "addendum" to my last post; i would highly recommend reading that one first.
going forward, i'll use addenda to """briefly""" expound on a side thread or theme from a prior post that i may find personally interesting, or that may have been controversial/lacking in overt evidence within the context of that post.
in my last bit of writing, i claimed:
...you can increasingly one-shot many popular SaaS companies in, say, cursor or claude code. a competitor has a feature you like? plug it into an LLM and ship it in a day... by and large, large language models are becoming increasingly generalizable for software solutions.
given this, one could argue that the SaaS industry is at risk of cannibalizing itself in the next couple of years due to AI-enabled oversaturation and overcompetition...
SaaS going under could lead to increased interest in funding biotech.
i wanted to expand a little more on my thought process here surrounding SaaS, and why SaaS losing favor as a VC darling could lead to scientific innovation.
simply put - i think venture capital is addicted to SaaS. there are a lot of overtly-spoken reasons for the surge in software investment, such as: "software and automation allow for 100x productivity." "more companies are looking for software solutions to remain competitive." "AI is the future of humanity." however, i think there's a big unspoken reason for the mass software investment, one that VCs are afraid to admit out loud - software is easy.
compare software to any other traditional industry. is there any other sector where you can, with nothing but a laptop, go from idea to $100K ARR in a month? the combination of insanely low input capital, insanely high revenue ceilings, and hyper-compressed timeframes is exactly what drew venture firms to software. VC is an industry where quick, high-value exits from minimal inputs is a wet dream of a success story - why on earth would you invest in anything else if software enables that exact exit over and over again?
VC's addiction to quick and easy software exits, however, comes at the detriment of every other venture sector. many solutions that will enable a better human existence are not quick, easy, and fast. we cannot rely on academia alone to de-risk these solutions; many academic breakthroughs are not optimized for mass use, and (as we've seen recently), funding for academic research is quite fragile. in our capitalistic system, we need private individuals and venture capital to chip in on "moonshots" too.
however, venture capital won't be able to rely on SaaS forever. the two factors that made SaaS defensible were easing friction, and differentiation - and both are under threat in ways that i don't think the industry can come back from.
in the pre-AI days, it was a lot easier to distinguish yourself based on these two points. the velocity of developing software was also a lot slower - even if you caught wind of a feature your competitors had and wanted to adopt it, it'd take a couple weeks to prototype, a couple more weeks to test and bugfix, and a few days to integrate with prod code correctly. better to spend that month plus developing code aligned with your own business thesis instead.
LLMs turned that idea entirely on its head. forget features - you can now clone an entire open-source project or startup with less than a hundred bucks of claude code credits in a day or less. compute and hosting have all gotten insanely cheap, so deployment is a non-issue. if you really wanted to, you could probably use AI to automate your customer outreach and service too - some companies already have.
right now, the reason this hasn't completely disintegrated the SaaS industry is that AI still can't do customer discovery - it doesn't understand pain points quite as deeply, it doesn't have access to internal complaints and industry-specific knowledge, and its knowledge can lag behind the present day even with internet integration. it also can't perfectly do competitive intelligence, so you have to maintain an active eye on your competitors to know what features to copy. right now, you get the benefits of AI velocity without all of the detractors of AI-enabled copying. but AI is getting better.
as large language models keep evolving, the barriers that hold them back from carrying out customer discovery and CI tasks will slowly dissolve. what results will be brutal. SaaS companies will integrate more and more of their competitors' features until it becomes impossible for VCs to distinguish who to invest in, and clients who to buy from; companies who normally would use SaaS products will leverage more powerful LLMs and cheaper compute to develop their own software, reducing the need for SaaS companies.
now as i mentioned in my last post, this scenario is, at a surface level, what the kids call "quite bad". a majority of venture capital money is currently invested in software and software solutions. if software customers decide they're better served vibe-coding their own solutions instead of flipping a coin on what indistinguishable SaaS company to buy from, whatever investments don't get wiped out will mainly be concentrated in LLM and software-hosting platforms like anthropic, google, and cloudflare. firms will have to largely pivot away from investing in "ai-enabled ___".
i, however, think this is a good thing.
removing software investment from the equation would force venture capital to do something that it, in my opinion, has not done in a while - take actual risks on novel science.
this is a long overdue course correction. during the eras we consider "renaissance" or "golden" ages of science, there was no quick and easy path for benefactors to see results out of their investment. benefactors had no choice but to make long bets. we've started to move back in this direction with private research institutes like the astera institute and renaissance philanthropy, but if SaaS ceases to be a quick and effective investment, the movement will expand a hundredfold.
biotech especially stands to benefit - not just because its problems are hard and its upsides compelling, but because of a specific combination of factors that make it uniquely positioned to absorb patient capital.
first, biotech's failure modes are expensive and slow, which historically made it unattractive to VC - but the science is getting faster and the tools are getting cheaper. what once required a decade and $100M is starting to require less of both. with increased capital inflow also comes increased opportunity to decrease these timescales and failure modes even further in ways that may not be as obvious in other sectors.
second, unlike energy or defense — the other obvious recipients of displaced VC capital — biotech has a clearly-emerging path to democratization on the scale of software. cloud labs, cheaper instrumentation, better computational tools, a growing community of young founders. displaced SaaS capital arriving into a biotech ecosystem that's already democratizing is a very different proposition than capital arriving into one that isn't.
third, the problems biotech solves don't go away when the economy softens. cancer doesn't care about interest rates. neurodegeneration doesn't respond to fed policy. the demand side is structurally insulated in ways SaaS never was.
the problem then, becomes: are we ready for this shift in investment? as i addressed in a prior post, i don't think we're quite there yet, but we could be quite easily. and that's good - because the window may be opening sooner than anyone expected.
till next time,
nb.