AI is not the new {#insert_word_here}
I know, you have seen them too. Lots and lots of fancy articles claiming that AI is the new whatever. The ultimate solution to all the problems a company may have, a miraculous technology to worship.
This usually happens whenever a new technology emerges and requires deep technical knowledge to understand what it is really going on. By the way, in this case, the technology is not even new.
What happens when you put together a promising technology with fancy marketing? You get campaigns targeting companies promising them that if the use their {#insert_name_here} AI solution or services, they can solve their {#insert_business_problem_here}. This kind of marketing hurts everyone. Hurts the seller itself, who is probably overselling their solutions, and hurts the buyer, who is losing a good opportunity to apply a really good solution and, more importantly, to think about what digital foundations is his company still missing to be really successful. I mentioned this in one of my previous articles ("Applying human intillegence to artificial intelligence"), and it has recently been covered by and HBR article stating that if your company isn't good at analytics, it's not ready for AI.
One of the latest AI is the {#insert_word_here} that I have seen is the one claiming that AI is the new UI. Nothing farther from truth. AI actually enables new kinds of UI, and that requires to put a lot of attention in user experience and user interaction involving AI. My colleague Jorge Márquez talks about it in his recent article.
Head of Data & Intelligence Practice at Marionete
7 年Words of wisdom David: on getting analytics right before any ML journey. So many times we go live with products or services without having a decent MI (Management Information) layer to support your basic questions or inisghts. Still a lot to do....
Technology Strategy | Enterprise Architecture | Platforms, Data and AI
7 年Well I wouldn't say that AI technology is not new; the concepts behind the current wave of deep learning solutions might be some 40-50 old, but only now computer hardware advances have allow the "democratization" of AI (using Google's words), and so the "boom" of people and business trying to apply it everywhere... of course with mixed results. It will take some time for the IT market to understand the real capabilities of these technologies and use them efficiently.