6 ms·

So yeah, Sam is trying to recruit people into starting these kinds of companies. Hard tech has always been "back" (when did it go away, Elon Musk?), but it is

So yeah, Sam is trying to recruit people into starting these kinds of companies. Hard tech has always been "back" (when did it go away, Elon Musk?), but it is quite the pivot from the YC of yesteryear that capitalized on things like better UX for AJAX calendars, web hosting, and drag-and-drop file storage. The really hard tech though really evokes that initial visceral reaction of "too risky" for investors, especially if there's no real indicators in the form of traction or an MVP ("well just wait a sec there, professor, the problem is hard, so we haven't solved it yet"). As an investor, I certainly wouldn't feel comfortable shoveling stacks over to some guys who told me they were going to build true AI with a decade-long outlook. Yikes! It's also a tougher proposition for founders. You're basically betting 5-10 years of your life on a problem that you don't even know you can solve with no revenue/exit strategy in sight. Meanwhile, that guaranteed salary at Google sure is looking more and more appealing. I would almost recommend graduate school for these types of people looking to leave their mark on the world in solving a really hard problem where any real contribution only inches the world closer to solving it. The struggle is real for all actors here. For founders, the trick is finding that sweet spot where a problem sounds hard on paper (such as self-driving cars or VR headsets, wooo!), but actually is feasible using current technology (e.g. stick some lenses in a piece of cardboard), but due to timing or market forces or whatever, nobody is currently paying attention to it yet. Then at least, you can execute just like an AJAX calendar app would and obtain the same outcome (to vastly understate the challenges involved!).

vannevar 11y ago
For founders, the trick is finding that sweet spot where a problem sounds hard on paper (such as self-driving cars or VR headsets, wooo!), but actually is feasible using current technology... This is how true innovation happens. Like evolution, innovation is not a steady inching forward (though that kind of progress certainly operates continuously in the background), but a series of sudden leaps punctuated by relative stability. Those sudden leaps happen when the right brain is in the right place at the right time to make the right analogy or synthesis with existing ideas. You can't fund people to produce it on command, you can only hope to be among the first to recognize that it has happened and to invest in it at that early point, before the innovation is disseminated. For an approach that does seek to produce innovation on command, see Ken Stanley's work on novelty search: https://www.youtube.com/watch?v=dXQPL9GooyI https://www.youtube.com/watch?v=dXQPL9GooyI
nostrademons 11y ago
AJAX calendar apps were themselves in that category when they came out. The conventional wisdom pre-2004 was that webapps could never replace the vast majority of desktop apps, because they required a hard page refresh and lots of clumsy session-management whenever the user performed an action. But a few technologies (XMLHttpRequest/IFRAMES, developed in the late 90s but not well publicized; faster computers & JS interpreters; and cross-browser JS libraries that abstracted away browser differences, developed in 2003-2006) had silently made it possible to build all your app logic in Javascript in the browser, and then just communicate with the server via a tether. This was not at all how people made websites in the early 2000s. I still remember seeing Outlook Web Access in 2001 and thinking "Wow, you can do this?" People think of AJAX calendars as a joke startup idea now because the techniques for doing this have been well-publicized for a decade now. The trick to creating (what appears to be) a revolutionary startup is to find other techniques that are not well-publicized and then apply them in a new, surprising, and useful domain.
- Hydraulix989 11y ago
  Then why make the distinction? Why are certain startups considered "hard"? Certainly, the real legwork was done by the browser authors at the time who revolutionized the web by adding these features? Is it more of a call-to-action? "Make startups!"
  - nostrademons 11y ago
    In some ways, I don't think it is a useful distinction. I had a boss once say (tongue-in-cheek), "It's work that I don't know how to do, which makes it hard." I think that's how most people actually think about it: "hard tech" is anything I don't know how to do, while "trivial problems" are anything that I know how to solve. There are certainly some things that other people view as hard - like searching the web on every keystroke - that I view as...well, not trivial, but a simple matter of applying capital and writing code, because I know how it all works behind the scenes. I'm guessing this article is attempting to focus attention away from all of the "The fundraising environment is falling!" and "My mobile app never got traction!" posts to where there are opportunities to exploit an information monopoly, and trying to recruit founders with this specialized technical information to YC. (As a side note, I actually expect we're in for a large mobile renaissance in a year or two, with many more actually useful mobile apps becoming commonplace. But the "specialized knowledge" that will drive this is market knowledge, not technical knowledge. It's people with intimate knowledge of a problem space being able to acquire just enough technical skills (or partner with a mobile app dev) so that they can build & market a useful app. This is a power shift from technical founders to business founders in the mobile space, while the technical founders go off and build hardware/VR/AI startups.)
ibejoeb 11y ago
>The really hard tech though really evokes that initial visceral reaction of "too risky" for investors, especially if there's no real indicators in the form of traction or an MVP I feel like there is probably near parity in the risk between these "hard tech" problems and things like social tech, etc., that many investors have gravitated toward. It's like "Can we build it?" vs "Will they use it if we build it?"
- Hydraulix989 11y ago
  Unfortunately, the investment climate these days focuses more towards the latter. "Make something people want" is what they say. I really do hope Sam's article raises attention more toward the former though. I suspect you'll see more startups coming out of that bucket in the impending winter while the latters' collective refrigerators all run out of Odwalla.
petra 11y ago
>> For founders, the trick is finding that sweet spot where a problem sounds hard on paper (such as self-driving cars or VR headsets, wooo!), but actually is feasible using current technology Or create short-term intermediate technologies and make money of them and use that for r&d on the hard part.
- Hydraulix989 11y ago
  It's pretty tough to do that though. You're essentially building O(n) businesses.
  - petra 11y ago
    You don't have to build O(n). Just build one business as a revenue source. For example theranos sold lab services for pharma companies, which had easier requirements than blood testing(which require thousands of tests, and other stuff). Google's Verily uses the same strategy - selling testing and data services for companies, while build it's future patient platform.
    - Hydraulix989 11y ago
      I'm not sure how citing the most controversial startup in the Valley (who allegedly raised boatloads of cash with a fundamentally broken product) and a big company with infinite cash as examples really helps your case? I couldn't even find anything about the "services" Theranos offered (and it seems unlikely they would pursue the multi-year bureaucratic hurdle-jumping FDA approval process for two different products concurrently). I only see something about an equally controversial SARS detector but with strong indications that it was developed pre-Theranos by the founder.

vannevar 11y ago

For founders, the trick is finding that sweet spot where a problem sounds hard on paper (such as self-driving cars or VR headsets, wooo!), but actually is feasible using current technology... This is how true innovation happens. Like evolution, innovation is not a steady inching forward (though that kind of progress certainly operates continuously in the background), but a series of sudden leaps punctuated by relative stability. Those sudden leaps happen when the right brain is in the right place at the right time to make the right analogy or synthesis with existing ideas. You can't fund people to produce it on command, you can only hope to be among the first to recognize that it has happened and to invest in it at that early point, before the innovation is disseminated. For an approach that does seek to produce innovation on command, see Ken Stanley's work on novelty search: https://www.youtube.com/watch?v=dXQPL9GooyI https://www.youtube.com/watch?v=dXQPL9GooyI

nostrademons 11y ago

AJAX calendar apps were themselves in that category when they came out. The conventional wisdom pre-2004 was that webapps could never replace the vast majority of desktop apps, because they required a hard page refresh and lots of clumsy session-management whenever the user performed an action. But a few technologies (XMLHttpRequest/IFRAMES, developed in the late 90s but not well publicized; faster computers & JS interpreters; and cross-browser JS libraries that abstracted away browser differences, developed in 2003-2006) had silently made it possible to build all your app logic in Javascript in the browser, and then just communicate with the server via a tether. This was not at all how people made websites in the early 2000s. I still remember seeing Outlook Web Access in 2001 and thinking "Wow, you can do this?" People think of AJAX calendars as a joke startup idea now because the techniques for doing this have been well-publicized for a decade now. The trick to creating (what appears to be) a revolutionary startup is to find other techniques that are not well-publicized and then apply them in a new, surprising, and useful domain.

Hydraulix989 11y ago
Then why make the distinction? Why are certain startups considered "hard"? Certainly, the real legwork was done by the browser authors at the time who revolutionized the web by adding these features? Is it more of a call-to-action? "Make startups!"
- nostrademons 11y ago
  In some ways, I don't think it is a useful distinction. I had a boss once say (tongue-in-cheek), "It's work that I don't know how to do, which makes it hard." I think that's how most people actually think about it: "hard tech" is anything I don't know how to do, while "trivial problems" are anything that I know how to solve. There are certainly some things that other people view as hard - like searching the web on every keystroke - that I view as...well, not trivial, but a simple matter of applying capital and writing code, because I know how it all works behind the scenes. I'm guessing this article is attempting to focus attention away from all of the "The fundraising environment is falling!" and "My mobile app never got traction!" posts to where there are opportunities to exploit an information monopoly, and trying to recruit founders with this specialized technical information to YC. (As a side note, I actually expect we're in for a large mobile renaissance in a year or two, with many more actually useful mobile apps becoming commonplace. But the "specialized knowledge" that will drive this is market knowledge, not technical knowledge. It's people with intimate knowledge of a problem space being able to acquire just enough technical skills (or partner with a mobile app dev) so that they can build & market a useful app. This is a power shift from technical founders to business founders in the mobile space, while the technical founders go off and build hardware/VR/AI startups.)

ibejoeb 11y ago

>The really hard tech though really evokes that initial visceral reaction of "too risky" for investors, especially if there's no real indicators in the form of traction or an MVP I feel like there is probably near parity in the risk between these "hard tech" problems and things like social tech, etc., that many investors have gravitated toward. It's like "Can we build it?" vs "Will they use it if we build it?"

Hydraulix989 11y ago
Unfortunately, the investment climate these days focuses more towards the latter. "Make something people want" is what they say. I really do hope Sam's article raises attention more toward the former though. I suspect you'll see more startups coming out of that bucket in the impending winter while the latters' collective refrigerators all run out of Odwalla.

petra 11y ago

>> For founders, the trick is finding that sweet spot where a problem sounds hard on paper (such as self-driving cars or VR headsets, wooo!), but actually is feasible using current technology Or create short-term intermediate technologies and make money of them and use that for r&d on the hard part.

Hydraulix989 11y ago
It's pretty tough to do that though. You're essentially building O(n) businesses.
- petra 11y ago
  You don't have to build O(n). Just build one business as a revenue source. For example theranos sold lab services for pharma companies, which had easier requirements than blood testing(which require thousands of tests, and other stuff). Google's Verily uses the same strategy - selling testing and data services for companies, while build it's future patient platform.
  - Hydraulix989 11y ago
    I'm not sure how citing the most controversial startup in the Valley (who allegedly raised boatloads of cash with a fundamentally broken product) and a big company with infinite cash as examples really helps your case? I couldn't even find anything about the "services" Theranos offered (and it seems unlikely they would pursue the multi-year bureaucratic hurdle-jumping FDA approval process for two different products concurrently). I only see something about an equally controversial SARS detector but with strong indications that it was developed pre-Theranos by the founder.