Loading...

L’ensemble des contenus Business Digest est exclusivement réservé à nos abonnés.
Nous vous remercions de ne pas les partager.

Book synthesis

AI is becoming more and more like us

AI is getting better at imitating powerful human cognitive traits. More human-like and more accessible, the technology is helping breathe new life into jobs, business models and potential innovations.

The convergence of speech and visual recognition technologies, natural language processing and deep learning is making AI smarter and more like us. It has been so successful that the head of AI at H&M prefers to talk in terms of “augmented intelligence”. So, now you have no more excuses for keeping AI at arm’s length! 

Based on

Based on Radically Humanby Paul Daugherty and H. James Wilson (Harvard Business Review Press, 2022). 


More human and less artificial   

Even though AI still has a long way to go, it is now capable of adopting ways of reasoning, operating, communicating and learning that are closer to our own, which opens up even more opportunities. 

  • AI talks like us. AI’s ability to use natural language (known as natural language processing) is growing exponentially, as can be seen in the success of GPT-3. This expands AI’s capabilities enormously: its level of understanding is much better (even though irony is still something of a problem!), it knows how to write creative text, and it can handle more language-related tasks. All this means it is easier for anyone to interact with AI sometimes without even being aware of it. 
  • It knows how to fire up its “neurons”. While the linear functioning of algorithms is still the norm, their neural functioning in the manner of the human brain is growing fast. What’s more, AI is learning not to process every bit of information it receives so that it can be more effective and consume less energy. 

How can Zappos help its customers find the shoes they want quickly when there are over 100,000 different models? The platform has introduced a Darwin-inspired search engine where a range of algorithms compete with each other: the one that delivers the most relevant result is “rewarded”, which serves to embed the learning. 

  • Learning to be autonomous. It is much more difficult to design fully autonomous automobiles than originally expected, and AI performs better when it interacts with humans than when operating autonomously. But the technology is also making progress in this area: Google’s stratospheric balloons, which are designed to provide internet access, can take their own decisions. Even more importantly, they can fine-tune them to suit the circumstances, which would have been out of the question only a short while ago. 

75%
SIGN IN
SUBSCRIBE TO
THE PUBLICATION
SUBSCRIBE TO THE PUBLICATION
See all subscription plans

© Copyright Business Digest - All rights reserved

Florence Meyer
Published by Florence Meyer
Executive coach, change management expert, and author. Constantly on the lookout for the latest management and leadership trends.