3 ms·
The problem is you're looking for easy stuff to do. Here's four ideas I just thought of: 1) Write software that uses machine learning to search for tornados in
by exabrial 9y ago
The problem is you're looking for easy stuff to do. Here's four ideas I just thought of:
1) Write software that uses machine learning to search for tornados in doppler radar signal returns
2) Write software that manages inpatient health care, orders, prescriptions, labs, etc
3) Write software that uses machine learning to route network packets faster
4) Use machine learning to control an engine's ECU to maximize fuel economy or power via valve timing, spark degrees/timing, fuel/air mixture, injection timing, compression ratio, etc
- lj3 9y agoThis stuff sounds more like R&D than markets to build products for. Way too risky for a company that hasn't even proven they can deliver a product where most of the technical problems have already been solved.
- exabrial 9y agoInteresting comment, I assumed all good idea required r&d!
- lj3 9y agoThat's the mindset of an engineer at work. Engineers want hard, sexy problems to chew on. Businessmen want stable, predictable tech with large margins (low production costs and a high selling price). This is why a lot of people talk about converting an excel spreadsheet into a SaaS business. It's ridiculously easy, low cost, stable, predictable revenue that companies will happily pay for to not have to deal with their excel spreadsheets ever again.
- exabrial 9y agoI fail to see how those are competing objectives...