Introducing: FakeAI companies

In recent days the network has been going crazy around a new study that found that 40% of European companies that say they use AI do not really do so. Wow.

The study looked at about 2,830 companies who defined themselves as companies using artificial intelligence in some way, but when MMC’s research team went in, they discovered that there was really no evidence of using a Machine learning or anything resembling AI. Why those companies did it? Most simply said it helped them raise more money from investors.

What do we learn from that? A. Not every company which adds “AI” to its name actually doing some AI work behind the scenes.

B. There is a need to make a very significant distinction between companies that use AI by developing algorithms independently, and companies that use ready-made packages or ready-made shelf products.

Do not get me wrong, there’s nothing wrong with the last option. On the contrary, I even encourage quite a few companies and organizations not to invent the wheel and use the variety of solutions we have today. Not everyone needs a monstrous algorithm like Google, Facebook, Amazon etc. We have made such tremendous progress that even “shelf products” can provide great value.

The last and perhaps most important lesson is that if we want or not, almost every company will become an AI company over the next decade — whether it wants to or not.

Not because to attract investors or to be sexy in the eyes of the public, but simply because it will become a standard to take the data you have and do something useful with it. The barriers for using a true AI will fell off and we will not make such a big deal of it.

Just as almost every company today is an “Internet company” — it is clear to us all that everything is running on the Internet, so we do not even mention it. This will be exactly the situation with AI in a not insignificant number of years and this is a natural passage of the technology world.