When is AI more than AI
The shimmering network above shows the complexity of networks and gears reflected in the morning sunlight. That reminded me of AI shimmering in a new era of IT… This follows the series When is AI not AI?, Stop AI scraping your Internet data, When is AI good AI?, and When is IT bad IT?.
Update: follow on last post of this series, Divergence of AI.
Update 2: more on back-dooring by Chinese mentioned in 2nd last paragraph from The Register.
A recent article in The Guardian showed the opportunities and perils of being at the point of AI, yet betrayed by the very tech that surrounds it. Here is the book on Amazon, the bowels of which let to uncovering “Y.S”. It’s insightful if a little scary glimpse backstage AI on the world stage — it shows how human faillability can produce unexpected outcomes.
Yet this is nothing new. Consider this TED talk for an insight into what’s really going on:
But that’s almost old school, as the real threat it globally distributed IoT systems microchips — low-level devices that sit in air con units, house internet connectors, cars and nuclear plants — that are often un-secured and therefore open to hacking. The fact they mostly come from China raises concerns you have to go to an Asian paper to hear about here. There’s neither politically correct nor neutral way to say it, but when we see China and Russia bullying Europe and the US, it’s hard not to pause.
Add to that the fact that serious hacking is documented to be run by Chinese nationals, last of which this scary near-miss here — in a nutshell, a so-called back-door was discovered by one volunteer on another volunteer’s code to have been hacked… and it was so widely distributed, it’d given malicious users access to most Linux servers, which as it happens run most the the Cloud — in other words, cyberwarfare is here and now!
… in other words, cyberwarfare is here and now!
What has this to do with AI, you might ask? Well AI is a house of cards built on IT systems that are definitley not secured. Please allow another metaphor, it really looks to be a colossus with feet of clay. AI is really more than AI, always, as it relies on so many other bits & pieces, that we’re “snap dab in the middle” of Donald Rumsfeld’s unknown unkowns territory.
Yet in my brief experience in AI for good for my non-profit (follow the latest in blog.zolnai.ca) shows me that AI works best with human input, which limits IT running rampant. But as Google head of DeepMind said below, “we’re at the beginning of the beginning of an AI revolution”. Stay tuned for more AI as I keep using it for good at my non-profit.
