Last week’s F8 event was one of those rare moments where I clearly sensed that I was witnessing something HUGE. This was going to be a day to remember. My laughter (Andy Samberg and Mark Zuckerberg) and sheer astonishment (Timeline & Info Graph apps) were contrasted by a vivid recollection of users already writhing in social-media indigestion from the cosmetic changes Facebook introduced just a few days ago. And sure enough, here was Zuck, once again, with child-like exuberance lobbing paradigm morphing UIs at us that were undoubtedly going to lead to massive brain hemorrhaging for some risk-prone users.

I saw all of this magic unfold and remember thinking: Zuck just pulled a play from the Steve Jobs handbook. This kid with the shy smile, the king of the social network, just showed us things we never knew we wanted, but now that they were being shown, well, let me just say we, I, can’t get them soon enough. (How did we ever live without this?)

Of course, I wouldn’t be writing about it here if there was no lesson to be gleaned for the AiCE project. And here it is, and it’s not just something for me, but I believe it is for all of us folks who think, write and/or teach about AI: Our development efforts will greatly benefit from the same Jobs/Zuck strategy. After all, the proverbial “they”, don’t really have the faintest idea of what they want AI to be, though “thanks” to pop culture they (think they) know what they don’t want (e.g., a T-1000 or SkyNet).

So what does all of this mean? I certainly don’t claim to have all the answers and am curious as to other AI researcher’s thoughts on the topic. But let me offer this much: For starters, the pioneers, in all aspects related to modern development of AI, need to adopt an entrepreneurial spirit, think like Steve and Zuck. Build the AI models we would want to see implemented. Not because they are popular, or are driven by focus groups. No. We should build the AI legal frameworks we believe are the best; the coolest. Build the AI entities and applications we believe are the most compelling. Never fall for the what-is-familiar/safe siren song. Challenge the existing paradigms. Ask thousands of “why not this way?” questions. Only in this way can we help build the best AI ecosystem.

Tags: AiCE