AI is Not a Singularity

AI is not a singular word. It is a plural. At Social Starts, the concept of flexible, learning, responsive and subtle networks is fundamental and touches every area we focus on; we don’t look for, think about, or fear AI as some sort of universal solution or challenge.

We don’t believe in an over-arching human-like AI. Rather, we see a host of special purpose artificial intelligences that can each deliver flexibility, sophistication, deal with complexity and uncertainty and ultimately generate serendipity and delight in its defined sphere. That is the way human intelligence works. An artist has a different truth than an actuary, even if they are presented with precisely the same circumstances and data. A “Red State” intelligence parses the world differently than a “Blue State” intelligence. Or, as Noam Chomsky has been fond of saying, Eskimos actually experience twenty different kinds of snow because their language describes twenty different kinds of snow. Human intelligence is diverse and produces a joyous noise of local realities rising toward the sun.

The fact that we so often think of AI as a singular word leads to our concerns about a singular all-seeing all-knowing digital Intelligence becoming Our Robot Overlord.

In reality, the Singularity with be a Plurality; not Skynet, but a swarm of local AIs, each with its own proximate truth, contending against one another like the Israeli Knesset. If there is an overarching “intelligence” in this universe, it will be a God utterly challenged finding anything universal from this scrum of perspectives.

We think, in fact, that the common usage should not be AI but AIs. Let’s stop talking about Artificial Intelligence and train ourselves to talk about and believe in Artificial Intelligences.

At Social Starts, we see AIs as an emerging construct essential to all our areas of investment focus:

WORK PLATFORMS. AIs are being built into customer service chatbots and becoming part of how employees are being recruited, evaluated, motivated and compensated. Our portfolio company FunnelBeam, for instance, uses AI techniques to determine when is the best time to communicate with sales prospects, using what channels and what kinds of language.

COMMERCE. AIs are part of how messaging is delivered into commerce sessions, what product choices customer receive. Commerce sessions are also being tied to concierge services that are driven by or augmented by tightly trained AIs. Our portfolio company Dopamine blends AI deep learning and state-of-the-art neuroscience techniques to induce predictable future behaviors in audiences. Another of our portfolio companies, Everywear, allows human fashion concierges to become super-powered via AI-driven product recommendations they can use to help customers shop most effectively.

CONTENT. Machine learning is becoming the editor of all our futures. We believe the combination of stronger AIs and deeper social analytics will combine as the driving force behind the future of content and, more importantly, the experiences that content can create. Our portfolio company Slash is an AI driven keyboard for mobile devices that makes suggestions about what you plan to do based on what you type.

IoT SOFTWARE STACK. The speed and complexity of the IoT demand AIs, lots of them. IoT security is impossible, except as a function of probability analysis, which ties to AIs. But think of the vast difference between the AIs needed for the industrial IoT, which we feel may be the biggest area of IoT value creation near term and IoT for baby monitors in a quiet nursery. Our portfolio company Aevena is developing a control program for drones that lets them learn about obstacles and inhabitants so they can self-navigate and recognize when something out-of the-normal is happening.

Beyond the focus on AIs in our current investment areas, we know that AIs will create new markets as they grow more numerous and capable. So, we continually look at emerging start-ups with AI technologies at their core, to see where, on the edge, AIs are starting to change the landscape. Some months back, that analysis was how we picked up on early signs of bots. Now, we are seeing AIs emerge in diagnostics and other health areas, in VR and AR experiences, and in the search for new ways to inform people without siloing them into the current problematic echo chambers.