Earlier this week, Google’s DeepMind team published a paper describing AlphaZero, a new generic reinforcement learning algorithm that has done some remarkable things. First, in about eight hours, it taught itself to beat AlphaGo, a human-trained AI system that beat the best human Go players in the world. It also taught itself chess and Shogi (known as Japanese chess) in about four hours and beat the best human-trained AI systems at those games.
How did AlphaZero teach itself? The rules of the games were programmed into the system. Then, AlphaZero started to play itself. The more it played itself, the better it became. Over a few hours, it learned to play the games better than the games had ever been played.
It is extremely important to understand that AlphaZero is not thinking like humans think, and it is not capable of general-knowledge decision-making. AlphaZero has performed brilliantly learning to play and win at what is known as a “perfect information” game. In a board game such as Go, chess, or Shogi, both players know all of the rules, can see the entire board, can see all of the game pieces, know the starting position of each piece, and know every move that has been made. This is dramatically different from game play with “incomplete information” such as bidding in a programmatic ad auction. It is also different from game play where players have “imperfect but complete information” such as poker, or contract bridge, or negotiating rates at the upfronts.
Treating Your AI Coworkers with Respect
To put AI to work for your brand, you must first set up a game you and your AI coworker can win. The questions you ask your AI coworker to answer need to be appropriate. You can’t ask an AI coworker a question like “How can we increase sales?” It will not understand. Assuming you have the required data for analysis, you can ask a question like “What’s the last day of the season we can sell widget X at full price in the Chicago metro?” You can even ask a follow-up, such as “Should we ship the widgets to doors in a warmer climate or just mark them down?” If you properly craft your questions, you are far more likely to obtain usable outcomes. And, to be clear, you need a high level of data maturity, data governance, and data hygiene to accomplish any of this.
We Don’t Have Access to AI Yet
Yes, you do. There are all kinds of open source projects you can access. Many paid services offer freemium models (where you can start your project for low or no cost and pay only for the increased services you need). You can rent time on any one of a dozen cognitive clouds. You can even call your favorite AI consultant and have them help you get into the game. Even if you can’t get your business unit into the game, you should take the time to learn everything you can about how machine learning and AI work, and then develop a personal thesis about when and how the technology will impact your business.
As for timing, it will not be long before AlphaZero (or something better) is commoditized and made available as a service. At that point, you’ll see share of basket analysis and media mix modeling routinely done at grade school bake sales.
Exo-digitally Enhanced Humans (People with AI Coworkers)
The human/machine partnership you need to forge is the next step in your personal evolution. You’ve already outsourced parts of your brain to your smartphone. It stores and processes your contact list, maps, knowledge base, etc. Now you will enhance your cognitive capabilities by partnering with an AI coworker. You just need to start thinking for two. What parts of a given problem are best solved by a machine, and what parts need to be solved by a human? Working together, the two of you will achieve greater results than either of you could achieve alone. I know this sounds hard. It can be tough learning to split your thinking in two. But don’t worry. There are many support and encounter groups that offer help and guidance. And if you need human/machine couples counseling, I’m available.
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it.