top of page

Top 5 Risks for AI Startups

We are in middle of AI revolution and plenty of AI startups are springing to life everyday. As a founder or investor, we are very much interested to keep these startups afloat. Let's see what are the risks which can kill these startups or shoot them into oblivion, to protect against them.

1. No cash for training further

Higher accuracy and faster inference times are two competitive advantages an AI startups can have. Both of these demand continuous refining of approach and subsequent training. These trainings need significant computational resources which cost money.

Every founder and investor should reserve 6 months worth of training money. If not, red alert should be sounded as you should know you are losing your competitive edge over others.

2. No time for research

As we saw above, continuous refining of approach is needed to maintain competitive advantage. There is tremendous AI research going on in world right now. Literally hundreds of great research papers are being published everyday. And if you are not focussing and progressing research on own and keeping tab on ongoing research, you will end up losing to someone who does.

Remember a new approach can dramatically reduce costs and increase accuracy overnight for your competitors.

Be that competitor yourself.

Make sure a certain ratio of your time is going for research.

3. Changing approach too frequently

Slightly contradicting previous point, research does not mean changing your approach every week. First, you should know limitations of your approach and how it can impact CX in present and future.

If you research carefully, you will see all breaking points beforehand, and in that case you must change something now.

Or if some fundamental shift happens in research, on the way, and then you must adapt.

Otherwise plan carefully and focus on one approach for some time. Research deeply to enhance its benefits. Focus on customer experience. A deeply mastered approach with superior customer experience will beat a better approach with inferior CX.

4. No money for AI talent

A great researcher is million times more productive than an average researcher. A great developer is 1000 times more productive than an average developer.

Your spending shows your priorities. I don't think I need say more.

5. No focus on customer experience

You may have most accurate model and fastest inference and still get zero traction. How can this be possible ? Because traction depends on how you treat customers and how you solve their problems. That is what brings revenue.

Technology is useless until it solves customers problems.

Hire people who know how to speedily design functional, beautiful interfaces for customers and your domain, or do it yourself.

Talk to customers. Talk to customers. Talk to customers.

All C-suite should talk to customer at least an hour daily.

If not, you are already dying a slow death.

Recent Posts

See All


bottom of page