Live data from Hacker News

Ask HN: Why are most US startups not solving novel problems?

news.ycombinator.com

31–40 of 44 posts

Re: Ask HN: Why are most US startups not solving novel problems?

#32
What novel problem would you want startup founders to solve? Are those problems solvable with current technologies? Is there a distribution model that'll work for that solution? Will the unit economics work? How big is the market for this novel problem? Is it worth spending time and money on this problem instead of a "generic product" that seems to be making more money?

Re: Ask HN: Why are most US startups not solving novel problems?

#33
I mean, what do you expect them to be working on, Nuclear Fusion? "Novel problems" make up the tiny minority of all work. The really hard problems are either already solved or lie beyond the reach of the average javascripter. And there's more, easier money to be made in the banal.

Re: Ask HN: Why are most US startups not solving novel problems?

#35

It is a lack of morals. Most founders of startups only want money. They do not care if they make other people's lives better or worse. This started with ethics laws. Ethics only means following the law and making money. Seeking the greatest good for the greatest number was aborted when ethics became mandatory. Ethics literally turns people into sociopaths. They are legally prevented from considering the welfare of ot…

Is morals different from ethics? What's the difference?

Re: Ask HN: Why are most US startups not solving novel problems?

#37
Conceptual impedance matching is the biggest problem. To use a personal example, I've got this silly idea, the BitGrid, which might (but probably not) prove to be the optimum way to get petaflop computing to the masses. It's based on some obviously bad design choices (when viewed from the FPGA world, for example), so that someone familiar with that world would quickly come to the conclusion that it's just not worth consideration.

Programming the thing requires a completely new set of tools, as there's no Von Neuman architecture to be found. It breaks my brain a little bit, every time I work on it. Completely writing new software isn't something we've had to do since the 1960s. We've always been able to bootstrap off of old ideas, and old code. Effectively, none of the tools normally used to develop software help.

The "impedance mismatch" of this thing is just too big for anyone to even think about making a startup around it.

Something that was a small improvement on FPGAs would be far easier to sell, but there's not much investment money in hardware, let alone ASICs.

A similar example would be the Jet Engine... nobody thought it could work, and it required a completely new way of doing things. Thus, nobody wanted to try to develop it.

Big changes require stable supplies of long term research and development funding. That's something the market just isn't equipped to handle, especially when driven to meet the quarter's numbers.

Re: Ask HN: Why are most US startups not solving novel problems?

#40
If you want to have a focus on solving novel problems then the business world isn't the best place in general as companies are too growth oriented to take the necessary risks. I would investigate the sources of the research papers startups base their products on.
Post reply on HN