Good piece, I basically agree. AI methods have largely stayed the same, certainly with some technical improvements, but have improved in application considerably due to new data, sensor and architectural environments. Have been examining some of the claims for the past couple of years.
This is hard to explain to the enterprise that wants some of that 'magic' now, and discovers that 80% of the work will be getting all the preparations in place to tie to the business process. Often a messy business. If done properly that can scale well. So there is big opportunity in data rich verticals.
I would also suggest you look closely at business process, and analytics that could improve it, before you codify it with AI. Quoting an executive from back then: " I don't want my old methods frozen in time ... ". The process of analytics and AI should ideally be delivered together. That is where our successes arose.
Inspired by: In CWorld: " ... A look at the re-emergence of A.I. and why the technology is poised to succeed given today's environment ... "
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment