In short, this article can be summarized: "VC used to fund R&D-heavy startups like Genentech and Intel, now it's used to make startups grow as fast as possible at any cost, just like cancer". Mostly agree with this.
And this might be a good thing. Lots of the modern VCs, some even by their own admission, don't have the acumen necessary to fund a Genentech or Intel.
So while I'm not happy with the current situation, I don't see the point of scolding VCs for NOT doing things they're not qualified to do.
Of course, the purpose of VC isn't to fund innovation so much as make money for the LP's, so this really speaks to "What marketing pitches work on LP's?" But at some point they'll get tired of funding Uber vs. Lyft wars and look for a way out of Hobbesian struggle.
Theranos is an example of the dangers of investors that lack the expertise necessary to evaluate "hard problem" startups.
There are at least 3 parties involved in every VC funded startup, the company, the investors and the customers.
Growing as fast a possbile and becoming a market leader benefits 1 perhaps 2 of those. I would argue usually benefits 1.
Having a finite VC-funded runway means eventually having to convert market share to profitability, such as by raising prices, and that’s not always a great result for customers.
In the end it's often the VCs that choose the market leader, not the customers. Competitors that enter the market without VC backing are forced to be profitable from the start and have no way to compete on quality or price. This is especially apparent with "gig economy" startups where offering subsidies/incentives to the service providers increases the quality (timeliness, availability, coverage) of the service for the customer.
In many ways it resembles a command economy, with a small number of VCs deciding what "the next big thing" is ("oh, looks like everybody wants scooters, let's pour all our capital into scooter companies").
The market thinks it needs more software when what it needs is good software in the right place. But then the belief that an industry needs more software creates a problem – a lack of software. This comes with the host of other software-specific intractables which spurs on the growth of software-focused labor.
But in the process of converting an industry into one that is software dependent, you can capture massive market-share while gaining slightly greater efficiencies on the services provided by that industry but then introducing a whole new host of software-specific problems. Which grows software's demand.
The thesis that payoffs for VCs are so substantial in software startups come specifically from the scaling of software across an industry capturing industry-specific assets rather than improving the productivity of that industry per se. If it were just about outsized productivity more specific bets that resulted in critical efficiency improvements would be more common. But instead we see broad-based bets that serve to get rid of minor consumer inconveniences, then adding more inconveniences to produce an incentive to raise consumer prices.
How is software rent-extraction, and why would inferential computing be an exception to this? Software generally reduces the cost of performing some task while increasing speed, accuracy, and efficiency.
> The market thinks it needs more software when what it needs is good software in the right place. But then the belief that an industry needs more software creates a problem – a lack of software. This comes with the host of other software-specific intractables which spurs on the growth of software-focused labor.
Successful companies are looking for ways to increase their margins, which software generally does for the reasons I listed above, not inventing ways to integrate software into their business that serves no purpose.
> But in the process of converting an industry into one that is software dependent, you can capture massive market-share while gaining slightly greater efficiencies on the services provided by that industry but then introducing a whole new host of software-specific problems. Which grows software's demand.
Software generally is able to produce massive efficiency gains, not minor ones.
> The thesis that payoffs for VCs are so substantial in software startups come specifically from the scaling of software across an industry capturing industry-specific assets rather than improving the productivity of that industry per se. If it were just about outsized productivity more specific bets that resulted in critical efficiency improvements would be more common. But instead we see broad-based bets that serve to get rid of minor consumer inconveniences, then adding more inconveniences to produce an incentive to raise consumer prices.
Again, I question why you're saying that software doesn't produce large productivity/efficiency gains, and would add that VCs are interested in software companies because of the high margins, short production cycles, and past successes with software companies (among other reasons). You seem to be focused on a few specific types of companies (I think social media apps would fit your description, though I'd be happy to get specific examples of what you have in mind) that are neither the norm for the software industry or VC ecosystem in my experience.
Why be fooled by that kind of journalism?
It used to exist to fund research-heavy startups, that maybe governments would otherwise have to fund but maybe where governments shouldn't be sticking their paws in (in democratic countries anyway). But now it has become an industry of its own. Just like finance at large.