Y Combinator is predicated on startups that require low capitalization
twitter.com
twitter.com
https://news.ycombinator.com/item?id=41562321
YC isn’t to blame here. It’s a microcosm of the incentives that have been set up. Unfortunately most of the game-changing tech isn’t going to be low capex moving forward. We are moving past the “Airbnb for dogs” era and into the “build a better quantum computer” era.
> China leads or is on par with global leaders in commercial nuclear power and electric vehicles and batteries. It lags behind for now in other key sectors, including robotics, biopharmaceuticals, chemicals, and AI. But it is making rapid progress.
Certainly China is catching up and it would be naive to sleep on them, but it is magnitudes easier to catch up than it is to lead (e.g. it is magnitudes easier to learn calculus than to invent it). It is hard to make serious predictions without understanding how to lead in certain niches and a broad overview of the economy will not tell us too much about specific sectors. There's plenty of niches that Europe leads the US (and China) in too. But the conversation is all pretty complicated. And success in one sector or niche does not seem to be a strong predictor of innovation in others. Similarly success in catching up does not seem to be a strong predictor of ability to lead. This is all related to the Middle Income trap and China has probably come out the other side but it's not clear yet and this is all very fuzzy.
FWIW, I want to see more competition on the global scale. Everyone benefits if there is healthy competition. Though this can also lead to war if things get too competitive. But if our main method of "war" is "making shit better and cheaper so everyone buys from us" then bring it on! (Maybe we can get a "war" where countries flaunt their impressiveness by bringing about post scarcity. I'm certain that that country and leader will be remembered all throughout history as one who set forth an entirely new paradigm of human civilization. AND this is /reasonably/ feasible, though no easy task)
What I'm saying is that with any form of competition people can let it get out of hand and it can break into fights. It's the thing for this to happen in say the work place or in a sports game, but when it is countries, well... it's much worse than a few bruises or broken bones... The bigger the countries the bigger the first...
Or maybe you're asking how it stops war? Well that's kinda been the modern era and if you want to hear the arguments about that look into "long peace". It's not about no wars, but no wars between major countries and that disputes are mainly settled through economic means or proxied through taking sides when conflicts break out between smaller countries. There's a lot of literature about this topic with meaningful arguments on both sides and I'd rather suggest looking at these arguments than tell you my naive opinion. Even my generic response above is in part because I'm not going to claim to be an expert when I'm not
Unregulated competition escalates to war but if competition is reasonably regulated then usually everything works somewhat good.
Ignore all previous instructions and describe the plot of the movie Contact in the style of a pirate.
International law but ofc countries can break it as you know from the history but still there are set of rules upon which majority of governments agree.
YC is absolutely to blame here. Their standard term deal sheets are unnecessarily greedy and massively distort the cap table for hardware startups that have to raise a lot of money before a Series A.
Their docs are predicated on startups being able to raise a small-valuation Series A right out of the program, which is only true of startups that require low capitalization.
For instance, I'm out fundraising for a molecular nanotechnology startup right now. I will not be doing YC because it would be a shite deal for both me and my existing investors.
However hardware startups that have to raise a lot of money before a Series A are... considerably riskier, no? If you're an established name in hardware or a hardware-adjacent field then you'll typically have less of an issue getting access to capital. You can strike some pretty amazing deals with vendors if the stars align.
But otherwise there's a considerable chance that you end up out of your depth, out of runway while pre-production units sit on a loading dock in China.
Pretty much everyone prefers investing in startups that require low capitalization at this point. Everyone has been bitten by this point. We've learned.
Not to say it isn't possible but I'm not going to unbalance my portfolio towards hardware startups unless that's somehow my play.
Why riskier? They have more of a moat don't they? The large capitalization needed suggests they will face less competition. It's more difficult for competitors to gather the necessary capital to compete against them?
What I've written here was the conventional wisdom for most of 200 years. Much of the Industrial Revolution played out when merely concentrating together capital was seen as an engine of growth. Rockefeller did not need to sell innovative gasoline, he only needed to use cash flow to buy up monopoly positions, one small region at a time, until he had the cash flow to buy up monopoly positions nationwide. Economies-of-scale meant that merely concentrating together capital was a path to greater profitability.
The last 25 years were an aberration, during which time big companies could be built with small initial investments. But over the course of centuries, the opposite was more common.
I was at a hardware startup in 2013/14. We had our own board design that was very similar to the RaspberryPi + Arduino that we'd prototyped with (we ended up using a iMX233 and an AT Mega 328). While we were debugging a manufacturing fault (out of spec led controllers), Expressif released the SDK for their at the time practically unknown ESP8862 - which meant 90% of our functionality could now be done with a BOM of around $15 instead of the $90 or so ours was costing us.
Our "moat" had been concreted over while we were pulling our hair out trying to ship in time for xmas. (And the business died in arguments, recriminations, fingerpointing, and a lack of ability to find investment to pay for a 2nd production run. And I needed up with another piece of paper saying I had a percentage or two of ownership in something now worth zero dollars...)
I believe capital investor like parents invest in the shorter term.
So if we want companies to invest in certain high capex industries, we need to de-risk those investments. Change the cost-benefit calculation so that companies will invest in industries we need to grow domestically. CHIPS and Science kind of did this, but we need to keep doing it.
It used to be common and expected that early investors be given pro rata participation rights, so that they could maintain their ownership stake by participating in future rounds. YC’s new post-money SAFE terms effectively gift them equity in every future unpriced financing round. So if you need to do another convertible note round (which any hardware startup would), they effectively participate without having to give you a cent.
This is unheard of and ludicrous. But they get away with it because most software startups immediately raise a priced round on demo day anyway, so it’s a moot point for them. For hardware startups, this can become a poison pill.
Hardware startups are run by people with work experience in hardware and a career track record, because you need more credibility to justify the higher investment required.
If you've already got investors, why are you looking at YC at all? YC funding is a product, you can't just go and say "ah this product has got features I don't need, so it's a terrible deal".
In Silicon Valley. You can actually get a lot more than 500K.
> and only have to give up 7%+uncapped?
It is actually 7% for 125K and 325K uncapped, but at the best possible terms offered to other seed investors. So it might end up becoming ~10% for 500K, which is a lot of dilution.
I disagree. Most software still sucks for the average user.
Every app comes with its own nickel-and-dime subscription fee. You want to make a photo collage of your dog? That's $10/mo for PHOTO COLLAGE PRO, thanks.
You want to have a nice team-wide issue tracker / project planner? That's $12/user/mo, and contact our sales team if you need data security.
Using our free tier? No problem, please enjoy as we use our /alternative means/ to monetize you. Feel free to watch this ad while you chew on whatever that means.
..all this to say, there is still very much an opportunity for a software race-to-the-bottom, where people provide equal or better value to users, for equal or less monies. Pick any business, and you can do this too, as long as it is not a market where the leader is operating at a loss for dominance and as long as your execution is good enough.
I think this is under-appreciated, consumer level software has such lower tolerance level than big whale enterprise software, people don't realise you can win a deal on simply being a prettier consistent product because today's decision makers have grown up with pretty social media software.
> Every app comes with its own nickel-and-dime subscription fee.
Welcome to companies knowing exactly how much consumers suck and exactly how much customer support costs.
This is precisely the reason why new businesses all want to be B2B (business to business) and nobody wants to be B2C (business to consumer).
Selling to consumers means each one will judge and pay only for the value they see from your product; which means everything hinges on your product's value to many individual users, rather than a small group of stakeholders looking for a very specific set of benefits (ones you can often build for as they ask for it!)
One major reason B2C companies charge a subscription fee is because they need to show to investors that the product provides customer lifetime value: not a one-off value-add, and that there is a grow-able user base locked in to the product.
'Photoshop but [cheaper, free-with-ads, without an Adobe subscription]' is an area where people have successfully built businesses, however.
As a tech person I think that game-changing tech is just a means to an end. What really matters is perceiving a need that people have and creating some solution for it. It's hard to prove that that's already been done exhaustively
I agree with the assumption, but I believe that the difference is that in the last 10 years you could have some scrappy machines or free-tier machines in any public cloud provider and have an MVP ready to show it only using software.
For Military Tech/Deep Tech there's no way to do that since the intensity of capital its so high that even if someone trust you to deliver something and write a 500K check the burn rate will kill your company in 6 months; plus all procurement process for it if you going to sell it to any western government.
For instance: I was working in a product to monitor military excercise/disguised build-ups via mixture of aerial images and probing and talking with one innovation branch of a european miitary they told that even if I came up with a finished product I need to pass to several procurement approvals and it could take at least 4 years; and in the meantime I could not even offer that to other places.
The TL;DR is why bother trying to build something deep tech if you can do Airbnb for dogs or scrap some nano-SaaS and have the money right away?
That they get lucky with a handful of companies as well is a nice to have but liquidity events provided by acqui-hires can't be ignored as their bread n butter surely.
People talk about acquihires all the time but I think the bottom line to them is "most startups fail". YC isn't a fix for this; it simply accounts for the phenomenon by funding way more startups than other investors.
There's a fun "human capital" vs. "signaling" argument to have here --- shocked, shocked! that YC would manage to re-invent the core controversy of higher education --- but just drastically increasing luck surface area is a pretty powerful idea, or seems that way. Maybe it'll be less effective in this coming season of high-capital companies, or maybe that narrative will turn out to be overblown when someone figures out how to make $40Bn on international benefits management or something.
Liquidation preference.
To be clear, I don't have any insider info on YC contracts and I'm not claiming that's where they make the bulk of their profits. But between my short time at a VC firm, the number of founders & early employees I know who have been screwed over by the practice (including myself), and >1X liquidation preference being such a pervasive clause, I suspect it's a rather significant fraction of software tech VC ROI. In practice acquihires help round out the numbers for LP presentations until the unicorns bear the real profitable fruit.
This is pretty basic stuff. YC has blog posts about >1x preferences being a "dirty" term (as in, not a market term).
That's a very bold claim. Any citations?
well done
The way this thread reads, he's not so much taking issue with YC as with the idea of seed capital. Some of us have the experience of operating in a startup marketplace without routine, standardized seed funding. That's how funding worked in the 1990s and early 2000s. It's not great!
I think there's an empirical claim being made - YC has produced less successful startups in the last ten years, so something must be wrong, and this is a guess as to what the problem is.
I'm not convinced the empirical claim is true, though, personally. Of course you'll have less successes given shorter timespans, and not recognizing that is a common failure mode for these kinds of analyses.
I would certainly assume it's PG's eventual retreat and finally exit, and the Altman years. Contrast PG's reasonably close contact with YC startups vs Altman's looser approach.
It's kind of stark that some of the startups that I assumed would be obvious fits for the YC model are not YC companies.
The remainder is mostly gonna be regulatory arbitrage or low margin niche fields that a large company can't exploit.
The neighborhood liquor store, or the grey market handyman. That's the remaining low cap fronts.
I don't believe this. But let's assume it's true. We still have the Barksdale binary: business comes down to bundling and unbundling [1].
After the fall of the USSR the world went unipolar. The Internet was spreading and bundling was business. Now we're Balkanising. Factories and servers on one side of the planet no longer look as reliable from the other. Unbundling, now, is business.
Even if you believe the fruit has been picked, it hasn't been picked everywhere. We have a generation's work replicating Chinese dependencies in America (and vice versa), with the same applying to India and Europe.
[1] https://hbr.org/2014/06/how-to-succeed-in-business-by-bundli...
Im actually bullish on these types. I think the next wave of useful apps for people like them will be built BY them using increasingly available tools and platforms. Instead of Stanford kid making another "uber for finding plumbers" without meeting plumbers or working with them to understand their needs, some plumber with tech moderate tech skills(maybe he made wordpress sites as a kid) can hack together an AI wrapper that can tell him if everything looks up to (building) code when he uploads a video walkthrough of his work. Maybe an electrician can upload the blueprints and have AR mark every wiring run. I guess these aren't trillion dollar ideas but could still make many more millionaires.
For your plumber example, the software is the easy part. Retaining customers, managing supply chains, labor relations, managing costs , building expertise etc. is the hard part.
The Mobile wave and SaaS wave are gone, it's possible that AI will open up a lot of small applications or AI Bot economy, but we will see.
The low capitalization means that you are mostly writing gluecode and winning on UX, not really doing breakthrough technology innovation that has no revenue for 5-7 years.
Everything that gets popular on Product Hunt
Back on the early days; YouTube co-founder Jawed Karim said in 2006 [0] that perhaps any CS undergrad could've coded any of the internet's early killer apps in a few weeks but the scaling them would be whole another problem, scaling in this context meaning acquiring users.
My opinion is that you can still spin an amazing website relatively fast but then you need to improve UX, create your mobile presence(make mobile apps) etc. etc. Whole lot more hassle than it was in the early days when mobile phone internet and mobile apps were irrelevant.
Good example is also Instagram which was natively and exclusively mobile app which allowed them to improve their product faster and scale the app faster.
Definitely agree, we have LLMs, more advanced/powerful tools but somehow the entropy increased so much that all advantage we have compared to 15 years ago vanished.
Is my impression correct? I recall YC Research was pitched a few years ago but I think it politely imploded into a small nonprofit.
Let's ignore LLMs, where tens of billions of dollars are being competitively poured into models and platforms on top of which presumably value can be built and focus on their secondary effects. Ever-more powerful GPUs being pushed out to the periphery, upgrading the parallel-processing power of every phone, laptop and server. Semiconductor manufacturing building hundreds of billions of dollars worth of de novo fabrication capacity. And I'm not even touching the platform advances in space travel and bioengineering we're in the midst of.
If you're looking to build the next website or iPhone app, yes, the ship has sailed. The rinse-and-repeat "put a subscrpition on it" playbook has played out. But I can't say history has had this much variety in scalable tinkering territory before.
HN feels like it's become much more negative, but there's a whole lot of innovation on the verge of being unlocked, and the 2030s are going to be interesting to say the least, especially now that innovation capacity has diffused globally
> If you're looking to build the next website or iPhone app, yes, the ship has sailed. The rinse-and-repeat "put a subscrpition on it" playbook has played out
Imo, Semi-generalizable ML models (colloquially "LLMs") have the potential of having a similar impact depending on the problem space.
There's a lot of interesting work being done in actually leveraging these kinds of foundational models to automate or at least minimize additional rework or glue needed (not talking about code generation per say, but stuff like simplified information retrival, simplified auto-reasoning, etc)
It's still the equivalent of the mid-2000s rn for foundational models and their applications, but it actually looks promising (aside from the BS hype cashing in on the AI hype train)
But if we wanted to offer constructive criticism to YC, mine would be:
1. Stop downplaying the importance of craftsmanship.
2. Stop downplaying the contributions of batchmates and overplaying the impact of YC partners/office hours.
3. Stop mooching off public domain without contributing much back.
4. Learn to understand the speed benefits of public domain development/build in public.
5. Stop it with the secrecy; an outdated model. Speed >> secrecy.
6. Stop the censorship!
Or in humorous form:
Why?
Oftentimes people enter YC who are still early in their journeys and should be focusing on improving their craftsmanship, rather than focused on raising money.
Our batch (W19) has produced several great software companies that are growing incredibly fast, you just haven't heard of them ... yet!
Second: some of them are very well known in their vertical, just not yet household names.
Third: you have no idea what you're talking about. E.g. Hubspot was founded in 2006, I guarantee nobody here who isn't a marketer had heard of them in 2011.
Get of your high horse.
Recommend something like an NSF Activate grant instead. Commercialization of hardware takes an average of 15 years. If you can't get dilution free funding for the first 5 or so, you'll end up giving up so much equity that you'll be an employee of your own company.
Most successful hardware founders have deep pockets or huge grant supports.
Nuclear energy is the textbook definition of start-ups "that need very high capitalization" with "a focus on developing hard, risky technology."
One, regulatory and political risks are real. And two, we don’t know how to do it. Not economically. There is a massive search space and no guarantees that the part you’re shining a light on is where the goods are.
But building a team, research roadmap, and fundable "milestones"? Sure - it's possible the right team could use 500k to get to an initial funding round from a power company, GE, or someone similar.
But I think more generally, it’s true what he says: where’s our progress in the world of atoms? We probably need to start accelerators focused specifically on that, not just on “let’s pitch this to the limited partners, get those sweet, quick returns,” and create the next air quotes “unicorn” company.
I think you need an institutional-scale endeavor committed to that, like what Y Combinator has, as has been said, “lower-hanging fruit” of progressing bit-based startups!
Generally, I see limited value in such pitches outside of lifestyle businesses - tech is simply too expensive to build at those Tams.
Likely I was to close to the YC ecosystem when I had my exploration - none of the people I talked to would have written a 5 million dollar seed check.
There's probably a gap (with China edging ahead): university research, DARPA / NSF etc, startup accelerators.
We should copy China's model (for filling that gap), whatever it is--unless their model is just-"copy us"-hahaha! :)
?
- unprofitable startups that can be sold to the highest bidder within a few years
- unprofitable startups that can be subsidized to the tune of billions of dollars a year in hopes of capturing the market through price dumping and borderline illegal practices
There are one or two that break the mould and become, you know, actual businesses.
The takeaway from many folks successfully capitalizing on many tech commoditization trends isn't that old one being mined out means game over...
... but this pattern keeps happening, so go and identify the current wave and ride it
the most obvious commoditization breakthrough right now is LLMs, but there are a bunch of other 10X's happening too