Isn't it risky to selectively back particular YC companies? There's always a negative signal when your lead investor doesn't participate in future rounds - if they start doing that more regularly with this new fund, then it might become a negative signal for a YC company NOT to have Continuity/YC as a follow-on investor (which sucks b/c the firms are so early stage).
Anyone know how VC firms that invest in all early-growth-late firms deal with this dilemma (realizing there are only a few cases)?
YC has explicitly stated they won't lead Seed or Series A rounds in order to avoid this conflict.
They are intentionally leaving economics on the table in order to better support their founders. At the stage Continuity invests ($15-$50M rounds), YC's early signal is far less important than the actual fundamentals of the business.
It's even better than that too, given their commitment to invest pro-rata. So the money is there if you can get another firm to lead. Because this policy is public, there shouldn't be signalling issues.
By doing this they don't have to consider if startups are suitable for follow up investors and thus avoid the Keynesian beauty contest problem of not selecting the "prettiest" startup: https://en.wikipedia.org/wiki/Keynesian_beauty_contest
> "It is not a case of choosing those [faces] that, to the best of one's judgment, are really the prettiest, nor even those that average opinion genuinely thinks the prettiest. We have reached the third degree where we devote our intelligences to anticipating what average opinion expects the average opinion to be. And there are some, I believe, who practice the fourth, fifth and higher degrees."
I hadn't heard of that before, but I'll keep it in mind for sure.
I was going to say that the obvious solution is to not play -- to buy and hold based on fundamental value, and specifically seek out promising but undervalued stocks -- but then, that solution doesn't offer the kind of growth that VC wants, or that a Keynsian beauty contest offers when you win.
I'm not sure why a pessimistic comment is at the top of practically every HackerNews article, but in this specific case, what data do you have to support YC becoming a "typical VC." What have they done in the past that has been typical? They have re-imagined the startup industry. Not in the way airlines say "we've re-imagined economy travel," I'm saying turned the startup world on it's head. With that said, YC having $1BB to incubate and then further grow companies sounds like a great fuckin idea. With an enormous fund that spans market caps, they'll be able to fund even more moonshot ideas, which is even more exciting when you're talking about companies that have double bottom lines.
Raising a billion dollars for a fund might be considered 'typical VC', or not, depending on one's perspective. Certainly the financial scale at which YC operates has changed over recent years. Practically speaking it is no longer ~$17k for ~7% to get through three months before demo day. While I believe that YC played a role in changing the economics of startups, the economics of startups have started looking more and more like "big money" investing...particularly as the idea of "startup" has morphed as the term has been adopted in popular culture and pension funds and insurance companies have gotten into the act.
At some point it will make sense to start thinking about YC as a staid financial institution (whether or not that point is today, is another question) rather than some plucky newcomer. The amount of money an organization has affects its operation. It's impossible to spend $1 billion carefully deliberating $20,000 decisions one at a time...and carefully deliberated $20,000 decisions one at a time was what made YC successful...and how and why it changed the startup landscape.
You are reading this too seriously and with many assumptions. "Please don't become X. I loved you as Y." is simply a statement of Wish and Preference. There is nothing in that comment this new effort by YC should/will fail, only that they prefer a different YC than one that could become a typical VC.
I think the value that YC provides has to do more with its people and ethos than it's platform. Although they were the first incubator, I don't think it's the incubator model that makes them who they are (as we've seen many other competitors with the same model pop up) but rather it's the people and ethos that its founders engrained into the system that makes Y Combinator what it is. For that reason, I'm inclined to believe that the more of the traditional fundraising process Y Combinator moves into, the better it is for the overall ecosystem.
What I got from reading Paul Graham's own writing on the genesis of YC, was that his key insight was that VC's should be funding more companies, and earlier, and giving founder's much more control.
The other key innovation was funding in batches, in classes. This created a close-knit ad hoc community with shared goals, and one in which teams whose ideas were not finding traction could join teams whose ideas were.
I don't think it's common for a VC to hand out free money. The best part is that with the data they get from these experiments, combined with the vast experience of starting and growing companies, they'll likely be able to find business models that are successful financially while simultaneously achieving the goals of what they are researching.
I'm no rocket surgeon, but off the top of my head, they could determine the personal financial implications of not having to worry about keeping a head over your kids' head and food on the table. If you give someone a basic income, what do they do with it? does someone use time to become more marketable and be able to earn more in the long-run instead of have to focus on a minimum wage job to feed their kids? Maybe there's a business model in using yourself as collateral. I've seen personal examples of this happening - high net worth individuals that meet exceptional people and give them a basic income so that they can focus on big ideas, not affording rent. Maybe through the data YC is collecting, they'll be able to find a scalable form of that.
I believe it's possible YC has overextended themselves such that top AI companies - the most valuable companies in the future - are going to first try Gradient Ventures or more specialized funds, and it may cost them their future. I understand they want to be a destination for AI, but I have a tough time imagining it. Maybe when they opened the floodgates to 100+ per cohort was when their brand became "the best anything and everything of startups," and if a rich, connected AI veteran launched his own AI fund, I'd be more excited to knock on that door first, if not an AI fund from an org intimately tied to a top AI shop.
> AI companies - the most valuable companies in the future
To me that's like going to a state fair 150 years ago, seeing the vendors of snake oil, and concluding that the most valuable companies in the future will be selling bottles of secret potions.
For "artificial intelligence", I see no actual intelligence and a lot of emphasis on artificial. The best of it is at its core some fairly traditional applied math and/or electronic engineering.
Instead, there're huge sections of the research libraries for pure math, applied math, physical science, electronic engineering, and computer engineering, and in comparison so far the best of AI is hardly more than a few books on one shelf.
How to do applied science, engineering, etc. hasn't changed; AI hasn't changed how; the only way to have
> AI companies - the most valuable companies in the future
is just to call all the most valuable companies, or safer yet, all new companies, AI. Silly talk.
A claim like
> AI companies - the most valuable companies in the future
will be part of what we've seen before: AI winter, AI spring of new hope, AI summer of scam, AI fall of failure, AI winter again. Get your warm clothing while it's cheap now because the next AI winter will be darned cold!
E.g., look up the little game of, say, matchsticks called Nim. Let AI learn to play it. Against just a common man in the street, it might do well. Against me, it will lose at least half the games if we first flip a coin to determine who moves first. Just why and how I'm winning, the AI program won't figure out; figuring out such things is so far just not what AI does.
Or, teach AI some high school plane geometry. Then consider the problem, given triangle ABC, by Euclidean construction find point D on side AB and point E on side BC so that the lengths AD = DE = EC. To do this, need to be a little creative; that is, need some actual intelligence. The people who taught the AI geometry program likely didn't teach how to do that, and, then, for the solution, the AI program won't know how to do it, won't figure it out, and won't do it.
Net, AI is not "intelligent". Chase down big trees in chess, checkers, Go, etc., play games against itself, accumulate a lot of statistics, and, then, play and often win -- okay. Be "intelligent"? Nope, not yet.
Or, one day I got via e-mail a specification of a problem in 0-1 integer linear programming. A first example had 40,000 constraints and 600,000 variables. Some people had tried "simulated annealing", ran for days, took the best results they had so far, results of unknown distance from optimality, and stopped. Drop that problem into a lot of integer linear programming (ILP) packages, and won't get very far. Write some software to do what such ILP packages do and call it AI, and again won't get very far. So, instead, need some new ideas, that is, from some actual intelligence. So, I put my feet up, popped open a cold can of diet soda, reviewed a little, thought a little, did some derivations in non-linear duality theory (early in that theory there are some darned cute results, e.g., get some surprising, useful convexity), designed and wrote some code, and in 905 seconds on a slow PC found a feasible solution guaranteed to be within 0.025% of optimality. Figuring out how to do that for that problem took some actual intelligence.
Here's one: I was in an AI group monitoring server farms and networks. So, suppose we are given instances of one or both of those and collect data, say, points, where each point consists of numerical data on each of several variables, say, 50, and where we get, say, 100 such points a second.
Now we want to define and detect in essentially real time anomalies as, really, a reason to start diagnosis and see if there is a real problem. For this, we want, and are quite serious about, false alarm rate: We want the rate adjustable in advance and, then, in practice to get that rate essentially exactly.
And when we get an anomaly, we want the lowest rate of false alarms at which the data would still be an anomaly.
Okay, AI/ML machine, how the heck to do that?
The answer needs some intelligence, e.g., borrow some from ergodic theory, get a finite algebraic group, sum over the group, etc. For that work, the current techniques of AI/ML won't even pass the giggle test.
Sure, taking my work, could program that and call the result "artificial intelligence", but really just programmed some pure/applied math. And, sure, will need an algorithm to make the computations fast, and that will need more genuine intelligence.
AI "intelligent"? LOL.
> AI companies - the most valuable companies in the future
> To me that's like going to a state fair 150 years ago, seeing the vendors of snake oil, and concluding that the most valuable companies in the future will be selling bottles of secret potions.
That actually would've been a good investment hypothesis though, and continues to be:
Fully agree with your point about AI though. People always gloss over the transition from "can be made good at things, with effort" to "general purpose intelligent" as if there's not actual magic required there.
Just guessing, but 'going global' may factor into this. Not as in more than one meeting center, rather, globalizing the marketing for startups. There is an interesting 'snowball effect' of successful exits growing the number of investors, and YC, even with the huge number of applicants it enjoys today, may be seeing quality diminish and therefore need to establish more permanent presence in other founder hot beds around the world. Perhaps that such a move is independent of this raise, but in order to invest the money, they'll need to keep the building the flow of recruits.
I am starting to wonder if they are falling into the
growth trap a lot of companies seem to fall. Instead of staying at a certain size and having a nice sustainable business, they grow too big and lose their character.
My guess is they will fund some Startup School founders with a small amount of money like they did with the Fellowship. Then again it's all up in the air. They raised 700M last year or so. They could also be piling up cash for bad times to keep YC companies private for as long as possible.
Readit News
Anyone know how VC firms that invest in all early-growth-late firms deal with this dilemma (realizing there are only a few cases)?
They are intentionally leaving economics on the table in order to better support their founders. At the stage Continuity invests ($15-$50M rounds), YC's early signal is far less important than the actual fundamentals of the business.
How does Axios know that they have done a deal if it hasn't been announced?
By doing this they don't have to consider if startups are suitable for follow up investors and thus avoid the Keynesian beauty contest problem of not selecting the "prettiest" startup: https://en.wikipedia.org/wiki/Keynesian_beauty_contest
> "It is not a case of choosing those [faces] that, to the best of one's judgment, are really the prettiest, nor even those that average opinion genuinely thinks the prettiest. We have reached the third degree where we devote our intelligences to anticipating what average opinion expects the average opinion to be. And there are some, I believe, who practice the fourth, fifth and higher degrees."
I was going to say that the obvious solution is to not play -- to buy and hold based on fundamental value, and specifically seek out promising but undervalued stocks -- but then, that solution doesn't offer the kind of growth that VC wants, or that a Keynsian beauty contest offers when you win.
interesting thought exercise to try to understand the error at each level.
At some point it will make sense to start thinking about YC as a staid financial institution (whether or not that point is today, is another question) rather than some plucky newcomer. The amount of money an organization has affects its operation. It's impossible to spend $1 billion carefully deliberating $20,000 decisions one at a time...and carefully deliberated $20,000 decisions one at a time was what made YC successful...and how and why it changed the startup landscape.
The other key innovation was funding in batches, in classes. This created a close-knit ad hoc community with shared goals, and one in which teams whose ideas were not finding traction could join teams whose ideas were.
I don't think it's common for a VC to hand out free money. The best part is that with the data they get from these experiments, combined with the vast experience of starting and growing companies, they'll likely be able to find business models that are successful financially while simultaneously achieving the goals of what they are researching.
I'm no rocket surgeon, but off the top of my head, they could determine the personal financial implications of not having to worry about keeping a head over your kids' head and food on the table. If you give someone a basic income, what do they do with it? does someone use time to become more marketable and be able to earn more in the long-run instead of have to focus on a minimum wage job to feed their kids? Maybe there's a business model in using yourself as collateral. I've seen personal examples of this happening - high net worth individuals that meet exceptional people and give them a basic income so that they can focus on big ideas, not affording rent. Maybe through the data YC is collecting, they'll be able to find a scalable form of that.
What is the purpose of combining the funds? The only thing I can think of is to obfuscate some of their losses.
Also, is the new round of funding because YC's assets are illiquid and they need a cash infusion?
To me that's like going to a state fair 150 years ago, seeing the vendors of snake oil, and concluding that the most valuable companies in the future will be selling bottles of secret potions.
For "artificial intelligence", I see no actual intelligence and a lot of emphasis on artificial. The best of it is at its core some fairly traditional applied math and/or electronic engineering.
Instead, there're huge sections of the research libraries for pure math, applied math, physical science, electronic engineering, and computer engineering, and in comparison so far the best of AI is hardly more than a few books on one shelf.
How to do applied science, engineering, etc. hasn't changed; AI hasn't changed how; the only way to have
> AI companies - the most valuable companies in the future
is just to call all the most valuable companies, or safer yet, all new companies, AI. Silly talk.
A claim like
> AI companies - the most valuable companies in the future
will be part of what we've seen before: AI winter, AI spring of new hope, AI summer of scam, AI fall of failure, AI winter again. Get your warm clothing while it's cheap now because the next AI winter will be darned cold!
E.g., look up the little game of, say, matchsticks called Nim. Let AI learn to play it. Against just a common man in the street, it might do well. Against me, it will lose at least half the games if we first flip a coin to determine who moves first. Just why and how I'm winning, the AI program won't figure out; figuring out such things is so far just not what AI does.
Or, teach AI some high school plane geometry. Then consider the problem, given triangle ABC, by Euclidean construction find point D on side AB and point E on side BC so that the lengths AD = DE = EC. To do this, need to be a little creative; that is, need some actual intelligence. The people who taught the AI geometry program likely didn't teach how to do that, and, then, for the solution, the AI program won't know how to do it, won't figure it out, and won't do it.
Net, AI is not "intelligent". Chase down big trees in chess, checkers, Go, etc., play games against itself, accumulate a lot of statistics, and, then, play and often win -- okay. Be "intelligent"? Nope, not yet.
Or, one day I got via e-mail a specification of a problem in 0-1 integer linear programming. A first example had 40,000 constraints and 600,000 variables. Some people had tried "simulated annealing", ran for days, took the best results they had so far, results of unknown distance from optimality, and stopped. Drop that problem into a lot of integer linear programming (ILP) packages, and won't get very far. Write some software to do what such ILP packages do and call it AI, and again won't get very far. So, instead, need some new ideas, that is, from some actual intelligence. So, I put my feet up, popped open a cold can of diet soda, reviewed a little, thought a little, did some derivations in non-linear duality theory (early in that theory there are some darned cute results, e.g., get some surprising, useful convexity), designed and wrote some code, and in 905 seconds on a slow PC found a feasible solution guaranteed to be within 0.025% of optimality. Figuring out how to do that for that problem took some actual intelligence.
Here's one: I was in an AI group monitoring server farms and networks. So, suppose we are given instances of one or both of those and collect data, say, points, where each point consists of numerical data on each of several variables, say, 50, and where we get, say, 100 such points a second.
Now we want to define and detect in essentially real time anomalies as, really, a reason to start diagnosis and see if there is a real problem. For this, we want, and are quite serious about, false alarm rate: We want the rate adjustable in advance and, then, in practice to get that rate essentially exactly.
And when we get an anomaly, we want the lowest rate of false alarms at which the data would still be an anomaly.
Okay, AI/ML machine, how the heck to do that?
The answer needs some intelligence, e.g., borrow some from ergodic theory, get a finite algebraic group, sum over the group, etc. For that work, the current techniques of AI/ML won't even pass the giggle test.
Sure, taking my work, could program that and call the result "artificial intelligence", but really just programmed some pure/applied math. And, sure, will need an algorithm to make the computations fast, and that will need more genuine intelligence.
AI "intelligent"? LOL.
> AI companies - the most valuable companies in the future
More LOL.
That actually would've been a good investment hypothesis though, and continues to be:
https://finance.yahoo.com/quote/ABC/
https://finance.yahoo.com/quote/CVS/
https://finance.yahoo.com/quote/Ko/
https://finance.yahoo.com/quote/BAS.DE/
https://finance.yahoo.com/quote/DOW/
Fully agree with your point about AI though. People always gloss over the transition from "can be made good at things, with effort" to "general purpose intelligent" as if there's not actual magic required there.
Deleted Comment
Deleted Comment
https://www.slideshare.net/mobile/dmc500hats/venture-capital...