Machine learning and AI underpin the future for Google, but is it the future we want?
Google is now a player in the hardware market, which is not all that special.
We are not talking about the release of a self-driving car, for which the tech community has been patiently waiting.
As Business Insider points out, the new Google products are simply aimed at competing with existing hardware from the likes of Apple, Amazon, Facebook, Sony, and Eero. So what’s special here? What unites these products and makes them uniquely Google?
“We believe the next big innovation is going to take place at the intersection between hardware and software with AI at the centre of it,” says Rick Osterloh, Google’s SVP and the man in charge of this new hardware push.
Machine learning and Artificial Intelligence underpin everything Google is doing, from the new hardware, to RankBrain, which is now the third most influential part of Google’s search algorithm.
This is a relatively new level of confidence in AI from Google. And investor confidence in Alphabet — Google’s parent company — is booming. As of October 5, 2016, the day after Google’s hardware release announcement, Alphabet stock was on the rise to nearly US$785 per share as of writing.
Potential investors and startups are well advised to continue watching Google stock, because its next moves will be an indicator of just how well a committed dalliance with AI can pay off.
Also Read: The Google Pixel phones were almost made by Huawei
Earlier this year, Google showed us what its AI can do in a highly publicized faceoff with Go genius Lee Sedol. AlphaGo (Google’s artificial Go player) won four out of five games, proving AI can out-think humanity at a game we invented.
This result could be sobering: Machine pupil beats teacher, signaling a dystopian future in which our lives are dominated by machines.
In terms of Google’s hardware and search algorithm, the AI thrives on our personal data, which could signal an end to privacy (that is, if real privacy exists now at all). AI and the Internet of Things, thrive on big data. Could 2016 mark the beginning of the end of our control over AI, and therefore, a loss of control over the data on which it thrives?
Or, does the emerging prevalence of AI indicate, more than anything else, the power of gamification?
Gamification
Normally, gamification is a motivation technique. Gamification as motivation has gotten flak from some pundits, but it also has captured the interest of app and web designers, marketers, educators, and workplace managers, among others. Proponents seek to ‘gamify’ experiences in order to entice users to participate.
The term ‘gamification’ can take on a broad meaning. In an Appnovation blog post, successful gamification involves “introducing game elements and mechanics to enhance a product, service, or process.”
Large companies like Nike employ motivational gamification with its product Nike+ — an ecosystem of apps that give the user virtual awards for achieving workout goals.
The language-learning app Duolingo presents checkpoints, or levels, that encourage the user to continue learning.
Square advocates for loyalty programs as one of the easiest ways to get repeat customers. Repeat customers who go in for the loyalty program game spend roughly 67 percent more than first-time customers.
But as philosopher and video game designer Ian Bogost points out, the above methods of motivating people are more manipulation than actual gamification.
The applications are ‘exploitationware’ that use points and levels — the elements that merely provide structure and measure progress within a system — and call them the most essential part of games. Real games tap into something deeper.
Deepmind and Gamification
There is no question that Alphabet’s AI has benefitted a great deal from studying video games and the way the human brain works with games.
To create its machine learning algorithm, DeepMind studied video games to “create a single program that taught itself how to play and win at 49 completely different Atari titles, with just raw pixels as input.”
DeepMind understands games are not just about the checkpoints and badges. “Key game mechanics are the operational parts of games,” writes pholosopher and video game designer Ian Bogost. For humans, these operational elements — such as sensorial and kinesthetic adaptation, as well as risk calculation — produce “interest, enlightenment, terror, fascination, hope, or any number of other sensations.”
DeepMind studies deep reinforcement learning, which is the brain’s process of taking information from the past and applying it to new situations. In essence, this is a study of the complex way in which our neurons work together to make decisions when a challenge arises.
The neurotransmitter Dopamine plays a big role in providing reinforcement, working alongside hierarchical sensory processing.
In short, Google’s AI is true gamification. It is taking into account the way games captivate and engage people, and applying the reality to enhance a process.
Also Read: Google’s former M&A man in Asia just launched a new VC fund
By working to understand the actual way our brains play games and master them, Google hopes to produce results that are human, yet beyond the scope of what a single human could do.
In game theory, there is the concept of perfect information. A chess player can know every move the other player has already made and use this information to decide on the next moves.
If Google’s algorithm has perfect information — it knows each search you have made, and knows which searches all other users have made — it then needs AI to act like the chess player.
Google Hardware and Gamification
The obvious conclusion for Google, then, is that DeepMind’s mastery of the game will help Google master the hardware market. Later, it will underpin Google’s self-driving cars. When you participate by buying products and making searches, you’re helping DeepMind play its game.
Users are generating more data, and thus, deeper learning. But this is not a zero sum game, right? Google would have the masses believe that when it wins, they win. DeepMind’s slogan is ‘Solve intelligence. Use it to make the world a better place.’
But this is an extremely risky game Google is playing. It is more than a game, it is our lives. The data maps out a picture of the person.
In the hands of Google, data equals capital. In the mind of the machine, then, people equal capital. Is that really the type of game we want to play?
—-
Image Credit: melpomen / 123RF Stock Photo
The post Is Google’s pursuit of AI good for humanity? appeared first on e27.