The Sun Tzu of AI
aka how to AI
Artificial Intelligence is the headline on everyone’s lips and it’s a massive opportunity. The popular picture though is that AIs will soon replace most tasks (read ‘our jobs’). But if that’s all you are hearing, you are missing the much bigger opportunity. AI is about much more than the economics of replacing people. It’s about leveraging new kinds of intelligence that literally think different, can see what we miss and do what we cannot!* All this in addition to simply automating what humans can already do spectacularly well.
The real game changer in AI is not machines that think like us but machines that think in ways we cannot conceive!
Companies who do not imbue artificial intelligence into every aspect of their business, so much so that a completely new entity emerges, will be left behind in 20–30 years. But here in lies the catch. How can a current enterprise, focused on running their everyday operations, build their own superpowered future, if it requires a complete rethink, rebuild and novel attack? The first step is to realize that three distinct levels at which you can AI:
Level 1: Replace
This is unfortunately where 90% of today’s AI discourse is: use AI to replace human labor. Starting with tasks that are easiest for machines, this short-term thinking is a simplistic understanding of AI. Its popular because it’s an easy extension of the gestalt that hurled the 21st century upon us: automation à la the industrial revolution. ‘Replace’ also has a clear, quantifiable and attractive metric associated with it: the difference between the cost of the people replaced and the cost of the AI. The massive drawbacks remain largely unseen: your company loses access to human intelligence — currently still the most fearsome technology in existence! Secondly, what’s easy for a machine and what’s easy for a human are two very different things. Contrary to popular notions, with AI, “low skilled labor” might be surprisingly harder to replace than Ivy-leagued school economists. So beware of such reductionist approaches.
Level 2: Aid
The second level of AI is to aid humans. AI here can help free human mind space for higher value work. The clear advantage here is that human employees are almost universally overworked. Top executives, especially, have difficulty finding bandwidth for looking ahead. AI can takeover non-creative tasks and decisions and leave your employees with the mind space to work on making business better. With more mind space an unseen opportunity grows — time for creativity and innovation. Your organization learns faster, health care costs drop dramatically and due to some ancient metric unquantified by the industrial revolution, productivity grows!
Level 3: Amplify
The real game changer and the biggest opportunity in AI is not machines that think like us or do what we do, but machines that think in ways we cannot (yet) conceive and do what we can not. In games like chess and Go, we’ve learned that AI eventually beats human, but no AI till date can beat synergistic human + AI combo. More importantly, humans haven’t stopped playing chess (with or without computers) because AIs have beaten us at it. Amplify beats replace any day. While this is hard for us postmodern people to see, our ancestors would have understood this rather intuitively. They continually experienced the power of a companion that could see more, think different and sense via unknown modes. They knew that paying attention to a dog or a horse in the wild could mean the difference between life and death. Playing AI at the Amplify level sets you well above the curve.
With every tiny step made in AI compounding possibilities for the future, companies that do not amplify their intelligences with diverse automated intelligences will be left behind. The best part is you get to choose how you AI. Replace might be easier to absorb and implement. But such a short-term view only gives you short-term advantages. Building your own superpowered future requires a complete rethink: an amplification.
*For an engaging exploration of these ideas we highly recommend Kevin Kelly’s article “The Myth of the Superhuman AI” .
