Ask HN: What problem in your industry is a potential startup? (2019 Version)
This has been done in 3 of the past 4 years (2015, 2016 and 2018).
It will be interesting to see how different ideas are no longer cared of
(
2 comments
| 7 points
| on HN
)
Something like Framework7 is doing a good job of taking on Bootstrap or Ionic, but there's plenty of room there for improvement.
There's plenty of unicorns who could be doing a much better job - CRM, recipes, dating sites. I'd love to use a CRM but most of them are too complex, and you don't want to trust anything using vtiger with your critical business data.
Gonna do a startup based around arguably the worst problem to have when trying to build a neural network from scratch;
Choosing the architecture of the net itself. It's super cumbersome, and you can often go days/weeks with crappy models by choosing poor architectures/objective functions.
Currently working on AI to build AI architecture recursively.