3 ms·

I would change topics as: Is creating startups outdated? …making PG essays outdated. I’m serious. 1980-2020 was the hype of an undertaped market that was perso

I would change topics as: Is creating startups outdated? …making PG essays outdated. I’m serious. 1980-2020 was the hype of an undertaped market that was personal computing. Anyone in a garage could do something revolutionary because anything was already better than nothing. Even if it didn’t respect GDPR and leaked PII accounts like a sieve. That’s always like that in the start of a Schumpeter economic cycle, same happened with cars in 1900. Now, we’re in a consolidation phase. Or are we still in an “invention” phase? AI maybe?

ramesh31 4y ago
>Or are we still in an “invention” phase? AI maybe? I think AI will ultimately be the driver of consolidation. The innovation in software of the last 30 years was, as you said, from the ability of anyone with a laptop to dream up a big new idea. But with state of the art AI models costing prohibitive millions to train and deploy, the power has returned to the hands of capital.
- marban 4y ago
  Zig When Others Zag.
- slyall 4y ago
  On the other hand if AI is just an API call away it might allow lots of new products. Lots of bigger companies are offering AI Pixie dust, but pretty soon it might be something that every solo SaaS can also do.
mamonster 4y ago
I think the more important question is long-term interest rates. 1980-2020 is just a massive drill down in interest rates, which basically means that for the last 40 years you were incentivized to progressively take on more and more risk and bet on growth. If interest rates go back down in the near future, then DCF math will keep startups cool. If not, and we transition to a regime where you have maybe 4% interest rates as a floor, then startups as a culture will become outdated.
- disgruntledphd2 4y ago
  Im not sure about that. Startups in a higher interest rate world just need to be more profitable ideas, and need less capital. I think that's probably a good thing, overall. But who knows right, we're in a very odd economy at the moment so we'll see I suppose.
fhd2 4y ago
I remember three events that created a massive amount of startups: Personal computing, the internet, and then mobile. What these have in common are new ways in which people use computers, combined with a large expansion in usage. AI by itself doesn't achieve that in my book, voice assistants maybe, but I'm sceptical that they don't just move usage around. Suppose that's why people invest into stuff like VR/AR and neural interfaces, hoping those are the next big things that massively expand computer usage. Though computers are already being used quite massively.
- mmahemoff 4y ago
  Interesting way to look at it, but I think you could say the same thing about the internet, i.e., people still interacted with it using a PC. So I could still see AI having a similar impact without changing the physical form factor. That said, it might also be the case that incumbents have more to benefit than startups, in a world that rewards access to large proprietary data sets.
- tuukkah 4y ago
  Massive move of usage to new services should be enough for a massive amount of startups that build those new services. No need for expansion, I think.
lbriner 4y ago
If we are in the consolidation phase, I wonder if people now understand the risks that software can create and will therefore push for much more professionalism in the industry. Currently, in most countries, literally anybody can write code for something and sell it without restriction. Most people don't like regulation but we can't keep on producing insecure junk without Governments starting to require some kind of formal certification. It is never perfect as anyone in a regulated industry knows, but it has to be better than nothing. For example, how many companies use applications written 30 years with no concern for security and are never forced to retire it (or prove that it is secure)? How many companies don't even have visibility of the 1000s of web servers that have been accumulated over the years of merges and divisions etc?
- pooper 4y ago
  > professionalism One short story about how I got a rude email one morning at work saying we need to change all our connection strings in production (internal applications) because machinename1234.corporation.lan doesn't work anymore so they moved it to a new machine machinename1235.corporation.lan I still don't know how someone at one of the top five Internet providers in the nation thought it was acceptable to do that instead of renaming the new machine to the old one. It isn't like we were pinning certificates or something fancy. I am pretty sure the certificates were self signed, which is ok for on prem. > How many companies don't even have visibility of the 1000s of web servers that have been accumulated over the years of merges and divisions etc? I wish we had thousands of web servers. At one place I worked at it was the same set of few windows server boxes for everyone and everyone keeps tripping on one another's toes leading people to say well QA and preproduction are broken it will be fine once we are on production.
- acdha 4y ago
  Forget 30 years ago, how many companies are running apps which are still being written that way? I think about this a lot in the context of open source. We’ve had a ton of great things happen from everyone being able to share code but there is a valid concern about vulnerabilities in code which has been orphaned or was written by someone who wasn’t familiar with secure coding practices. For example, businesses have seen billions of dollars in benefit from WordPress based on the cost vs. functionality / productivity of commercial CMS applications but WP plugins are a famous source of exploits and many of them are written by people who don’t even have the time or skills to secure them if they did. It feels like there should be some way to change that but it’s not clear whether any of the efforts to fund open source will get enough traction.
hef19898 4y ago
Through one of my acquitances in the start-up world in Germany, I learned that even big tech investors (from the US), are now saying that the time of "big vision" is over and their portfolio companies should look to increase revenue before looking at developing the next suoer cloud platfrom solution for their hardware products. Sounds like solid advice to me.

ramesh31 4y ago

>Or are we still in an “invention” phase? AI maybe? I think AI will ultimately be the driver of consolidation. The innovation in software of the last 30 years was, as you said, from the ability of anyone with a laptop to dream up a big new idea. But with state of the art AI models costing prohibitive millions to train and deploy, the power has returned to the hands of capital.

marban 4y ago
Zig When Others Zag.
slyall 4y ago
On the other hand if AI is just an API call away it might allow lots of new products. Lots of bigger companies are offering AI Pixie dust, but pretty soon it might be something that every solo SaaS can also do.

mamonster 4y ago

I think the more important question is long-term interest rates. 1980-2020 is just a massive drill down in interest rates, which basically means that for the last 40 years you were incentivized to progressively take on more and more risk and bet on growth. If interest rates go back down in the near future, then DCF math will keep startups cool. If not, and we transition to a regime where you have maybe 4% interest rates as a floor, then startups as a culture will become outdated.

disgruntledphd2 4y ago
Im not sure about that. Startups in a higher interest rate world just need to be more profitable ideas, and need less capital. I think that's probably a good thing, overall. But who knows right, we're in a very odd economy at the moment so we'll see I suppose.

fhd2 4y ago

I remember three events that created a massive amount of startups: Personal computing, the internet, and then mobile. What these have in common are new ways in which people use computers, combined with a large expansion in usage. AI by itself doesn't achieve that in my book, voice assistants maybe, but I'm sceptical that they don't just move usage around. Suppose that's why people invest into stuff like VR/AR and neural interfaces, hoping those are the next big things that massively expand computer usage. Though computers are already being used quite massively.

mmahemoff 4y ago
Interesting way to look at it, but I think you could say the same thing about the internet, i.e., people still interacted with it using a PC. So I could still see AI having a similar impact without changing the physical form factor. That said, it might also be the case that incumbents have more to benefit than startups, in a world that rewards access to large proprietary data sets.
tuukkah 4y ago
Massive move of usage to new services should be enough for a massive amount of startups that build those new services. No need for expansion, I think.

lbriner 4y ago

If we are in the consolidation phase, I wonder if people now understand the risks that software can create and will therefore push for much more professionalism in the industry. Currently, in most countries, literally anybody can write code for something and sell it without restriction. Most people don't like regulation but we can't keep on producing insecure junk without Governments starting to require some kind of formal certification. It is never perfect as anyone in a regulated industry knows, but it has to be better than nothing. For example, how many companies use applications written 30 years with no concern for security and are never forced to retire it (or prove that it is secure)? How many companies don't even have visibility of the 1000s of web servers that have been accumulated over the years of merges and divisions etc?

pooper 4y ago
> professionalism One short story about how I got a rude email one morning at work saying we need to change all our connection strings in production (internal applications) because machinename1234.corporation.lan doesn't work anymore so they moved it to a new machine machinename1235.corporation.lan I still don't know how someone at one of the top five Internet providers in the nation thought it was acceptable to do that instead of renaming the new machine to the old one. It isn't like we were pinning certificates or something fancy. I am pretty sure the certificates were self signed, which is ok for on prem. > How many companies don't even have visibility of the 1000s of web servers that have been accumulated over the years of merges and divisions etc? I wish we had thousands of web servers. At one place I worked at it was the same set of few windows server boxes for everyone and everyone keeps tripping on one another's toes leading people to say well QA and preproduction are broken it will be fine once we are on production.
acdha 4y ago
Forget 30 years ago, how many companies are running apps which are still being written that way? I think about this a lot in the context of open source. We’ve had a ton of great things happen from everyone being able to share code but there is a valid concern about vulnerabilities in code which has been orphaned or was written by someone who wasn’t familiar with secure coding practices. For example, businesses have seen billions of dollars in benefit from WordPress based on the cost vs. functionality / productivity of commercial CMS applications but WP plugins are a famous source of exploits and many of them are written by people who don’t even have the time or skills to secure them if they did. It feels like there should be some way to change that but it’s not clear whether any of the efforts to fund open source will get enough traction.

hef19898 4y ago

Through one of my acquitances in the start-up world in Germany, I learned that even big tech investors (from the US), are now saying that the time of "big vision" is over and their portfolio companies should look to increase revenue before looking at developing the next suoer cloud platfrom solution for their hardware products. Sounds like solid advice to me.