Notre objectif est de mettre en partage sur nos trois spécialisations (stratégies et management de l'innovation business tous secteurs, stratégies de croissance ENERGIE et CLEANTECH, stratégies de croissance DIGITAL),
les analyses d'Innhotep, celles de nos invités et des articles tiers issus de notre veille.
Accélérateur d' "innovations business", Innhotep intervient comme conseil auprès de grands groupes et accompagne le développement de start-up high-tech.
But with its last three acquisitions — Boston Dynamics, Nest and DeepMind — it seems like Google is rapidly collecting the individual pieces to put together a “real life Internet,” a network of AI-driven robots and objects that could improve transportation, manufacturing and even day-to-day consumer life.
Google’s “real life Internet,” a business that reaches far beyond web search and online advertising, may look like a General Electric on the Internet of Things side, and an IBM on the software side — where artificial intelligence is at the core of products likeWatson.
At least that’s what it looks like right now, as the search giant is gobbling up almost every company that could fit into the puzzle, combining hardware, software, analytics, robotics and artificial intelligence into, well, something.
Google X, the company’s skunkworks unit that’s been developing driverless cars among several other sci-fi-esque projects, now seems to be leading Google’s hefty meatspace ambitions.
One obvious extrapolation from all these acquisitions is that Google will be in the business of data for a long time. Covering computers, tablets and now phones with Android and building applications like Maps to harvest information about its hundreds of millions of users, Google is now looking far beyond traditional computing devices. Acquiring Nest, which builds smart home devices, was one swift lunge in that direction.
How many more of these diversity acquisitions will we see before 2014 closes out?
Google is betting its future on the fact that one day our cars, refrigerators, mobile phones, computers and home devices will communicate with each other, generating insights that can be converted into data. And that these newer channels will result in a massive advertising opportunity.
But what can Google accomplish that IBM and GE cannot?
IBM has invested $1 billion in its AI-driven Watson project, which is expected to bring $10 billion in revenue over the next few years. Facebook too, has set up an artificial intelligence team to understand emotions, and according to The Information and a tipster, was even in the race to acquire DeepMind (our tipster held the Facebook bid at $450 million).
So far, IBM has depended heavily (perhaps doggedly) on Watson for making its artificial intelligence push work. Since its launch around three years ago, IBM has been pushing aggressively to turn its “Jeopardy”-winning computer into a business where healthcare and telecom companies pay to use Watson in real life. But as a WSJ piece earlier this month pointed out, IBM has been struggling to make it work.
On the enterprise side, both IBM and GE are still far away from making any big impact in terms of revenues, despite having the experience of working with Fortune 500 companies for decades.
Watson’s biggest challenge today is solving real-life problems and living up to the “intelligence” part of the artificial intelligence equation.
When asked by the New York Times what he wanted to build at Google, Andy Rubin brought up the example of a windshield wiper that turned itself on when it rains.
As humble as that sounds, Google ostensibly has a head start in terms of AI-practicality, with Google Now making strides in the proactive computing field. It also has a tremendous advantage in its treasure chest of user data, allowing it to predict and analyze patterns in behavior and needs more robustly than any competitor.
With one of the largest server architectures on the Internet, Google has the big computing power necessary for AI processing at its fingertips. It also has ancillary Google X efforts like Project Loon that could blanket areas in connectivity needed to power robotics.
A “real life Internet” may be closer than we think.