Enterprise AI: fiction and reality
If you are starting to suspect that the AI hype is overblown, and that it’s just the latest term that marketing folk love to add to make their product sound more interesting, you are not alone. You’d be wrong though – the AI effect is very real, and is set to make serious waves across many industries.
That’s one of the conclusions of a report I co-authored recently, but it comes with caveats. A big one is to understand that not all AI is equal, and that when the marketeers say ‘AI’ in relation to a serious offering, what they usually mean is ‘deep learning’, which is just part – albeit an important part – of the overall AI picture.
Even with this principle in mind, terminology can still trip you up. Where the practitioners say ‘deep learning’, many researchers prefer ‘deep neural networks’. it’s also important to realise that deep learning (DL) is not ‘intelligence’ as most people understand it. Rather than the clever androids of Humans and other popular science fiction, DL is the super-savant of AI: often able to out-perform a human expert, but only in a narrow domain.
Deep learning is AI at work
Still, what if that narrow domain is the quality control system on your production line, or the chatbot answering standard enquiries on your website? In that case, DL could be just what you’re looking for, both to improve response times and to offload the dull and repetitive stuff from your skilled people.
So while it is true that AI has been over-hyped, and that it was about time for a negative correction, there’s no need to fall into a trough of disillusionment. Approach it realistically, with an understanding that AI is not magic and has its limits, and it could pay off handsomely.
To learn more about Deep Learning, what it can offer and how to plan for implementation, you can download our free report here.