We're at the precipice of a transformation of the valuation framework we've been using for venture-backed software for over 20 years. We are going to need to revisit all stages of valuation frameworks and the broader Pre-seed to Series A,B,C... treadmill. The implications for every stage of venture capital (and perhaps software investing) will be jarring.

The advent of CodeGen (LLMs for writing software) and resulting transformation of software development will transform the economics of venture-backed software dramatically. In a world with 100x the number of developers who are 10x more productive, we'll see 1000x more software which will lead to increased software competition, lower margins & lower prices.

There are a number of phenomena at play here, all with different effects on the venture capital model:

1. The average hyperscale startup will require significantly less capital to achieve similar scale to predecessors. We will see multiple 10 person startups hit 100M+ in revenue.
2. We will need new valuation frameworks for tiny hyperscale companies. There is implicitly more risk in a world with infinite software. Startups are more vulnerable - more developers means increased competition, fewer employees means a larger % change in productivity / growth with every team departure.
3. Smaller, more productive teams will mean more profitability in some senses (smaller teams == less cost to deliver), and less in others (more software == competition which compresses margins).
4. With significantly fewer people capable of achieving hyperscale, the average startup will not need to raise nearly as much growth capital, resulting in far less dilution for early-stage stakeholders. 50% less dilution means a $500M exit results in the same returns as a $1B exit (and perhaps faster, which would enhance IRR)
5. Will IPO markets want to take risk on this new type of tiny software upstart where anyone leaving means possible imminent failure? Without more predictable exit opportunities and valuation frameworks, upstream capital will tighten and early-stage valuations will compress.
6. More focus on liquidity and profits; Given the higher inherent volatility of tiny hyperscale startups, investors will put more premium on extracting profits and less on growth (given failure rates will be higher);
7. The Death of Growth Equity? Will most software startups even need the $100M+ equity investment anymore when they have less than 20 people? Growth Equity will largely be secondary buyouts, but will revert to Discounted Cashflow modeling. History of cashflows, profits and expanding margins (which we've already established will be challenging in a world of 1000x software) will be critical to drive these types of liquidity events for teams;

In the end, it does feel like pre-seed and seed checks ($1-5M, enough to give small teams a couple years to experiment to find hyperscale) are of increasing importance in a world with 1000x software. We'll see more money flood to accelerators, and the increased competition will mean higher sustaining prices for early-stage investments. The questions that remain are largely around the distribution of outcomes in 2035 and beyond, notably:

1. With 1000x software, will we see 1000x the number of exits?
2. Who or what will be driving liquidity in a world with 1000x software? Public markets? Private secondaries? Some type of profit distribution framework?
3. And what will the most important valuation metrics look like for eventual liquidity? (profits? historical cash flow? growth? something else?)