One of the many announcements at last Apple WWDC was around a home speaker that looks very similar to Amazon Echo or Google home. Shape apart, it’s been positioned as just a music speaker with limited AI functionalities. This has disappointed many people, but there is a reason for it… it’s all about data.
Aratificial Intelligence and Machine Learning need data
Google, Amazon and Microsoft are furiously competing to take the market lead on the next wave of compute. Google is buying AI startups like crazy and has great AI-based applications, Amazon has been developing an entire ecosystem around its Echo, and Microsoft is putting Cortana on every Windows capable device of this world. They are clearly leading the pack, leaving a struggling Apple behind.
It’s not a matter of money, it can’t be said that Apple doesn’t have the power to invest in new technology or doesn’t have the right resources for it. And it’s not a matter of expertise, Apple has access to the same research as well as the same engineers in the market.
Apple doesn’t have the data.
We live in a data-centric era
Everything can produce data today, even your scales or any other device that is now “smart”. The process of replacing dumb devices with smart ones is going on in every home, workplace or any other place. Sometimes it’s not about the real smartness of the device itself, but just the fact that it can collect and send data to a central repository.
Moreover, sending data to the cloud is just the first step – to make it useful you have to record, store and analyze it. In fact, the very basic concepts about machine learning are concentrated around pattern recognition and computational statistics.
As far as we all know, Apple has strict policies about the data it collects, saves or analyzes. It puts personal privacy first limiting the amount of information available to tools like Siri. If you think about Google for example, it has the opposite approach – they analyze everything you do, all the time. Google scans your emails and searches, tracks your position and it does many other activities putting all of this in correlation with similar profiles while building a strong data set where tools like Assistant can find the patterns it needs to predict your next request, placing it exactly where you need it. And I don’t think Amazon or Microsoft have any scruples in doing that… They are like big brothers, and this is why.
Scary but useful
People prefer Google Assistant or Amazon Echo over Siri just because they are more useful – they don’t know, or they prefer to ignore, privacy issues (and, unfortunately, I have to say I’m one of them).
It is true that many Apple customers would trade some of their supposed privacy for a better service or a useful Siri. But, at the moment, it is also true that there are other features that compensate for Siri… and you can’t have everything. The problem for Apple is how long will it last, and will it be ready (or in the right position) to change it?
Closing the circle
This post is not all about Apple and its product strategy. My point is that next generation applications will need huge amounts of data and the right tools to make them effective. We think about machine learning and artificial intelligence as something that only large hyperscalers can do, and it is quite true now, but it will change soon.
More and more applications and infrastructures will take advantage of the technology pioneered by these companies and they’ll be soon available to a much broader audience (an example from my company here).
The challenge remains the same, as it happened in the past, it will be necessary to look forward and think about investing in next generation tools, infrastructures (and new processes) to take advantage of this technology before others. It’s either this, or it will be harder and harder to play catch up given the acceleration we are experiencing in the IT field.