What’s really happening with machine learning?

Artificial intelligence will be explored at this year’s SXSW Interactive festival

If artificial intelligence (AI) were a band, it would be the Beatles. The hype surrounding the technology in these formative years for artificial intelligence goes from tween-obsessed adoration to uber-conservative, album-burning vitriol. If we are to believe the scaremongering, however, and robots do develop the ability to think for themselves, they will likely not overthrow human civilization as suggested, but will choose instead to retreat into the lab, much like The Beatles did into the recording studio, to escape the hype and make beautiful, robotic art in peace.

In reality, AI hasn’t come all that far. There are so many “what ifs” being bandied about the artificial part of the narrative, it is gaining a life of its own.

I get it. When the alternatives for tech journalists to write about include data storage, fintech and medical diagnostics, AI offers sexier and more controversial tangents to steer one’s overactive imagination towards without fear of retribution for engaging in irresponsible or inaccurate journalism.

Perhaps tech writers get away with wildly hypothesizing because many readers don’t feel confident enough in their own understanding of certain areas of science and engineering to challenge them. I say this because most tech journalists – myself included – don’t always understand the stuff we are talking about either.


Utopian state

Could AI become smarter than its creators and enslave us? It "could". Or it might it bring us to a utopian state where all things are possible for all people. Maybe. War between Britain and Ireland could also break out post-Brexit but you don't read much about that 'what if' in the political pages.

The problem doesn’t begin and end with journalists. After covering numerous tech conferences such as SXSW Interactive over the past decade, I have come to realise just how many of the people giving talks about science and technology-related topics are neither scientists nor engineers. They are PR, marketing and communications gurus who either crave attention or have a commercial interest in the public’s perception of whatever gadget they are shifting this week.

A cursory glance at the programme for the SXSW Interactive festival, which is taking place in Austin, Texas, this month, provides enough ambiguity as to the true significance of machine learning to make anyone with no vested interest potentially love, hate, fear, marvel and laugh at AI simultaneously.

There are talks programmed with enough scope to show all sides of the debate, an economy of scale enjoyed only by festivals as large and well curated as Austin’s annual art, music and tech extravaganza.

The organisers go to great lengths to provide a variety of perspectives on hot topics in the realms of science, innovation, enterprise etc. I will go to as many of them as my "real" intelligence can handle. But the one I'm most excited by currently is a panel discussion entitled: AI – Actually Still Terrible.

Pretty dumb

As the SXSW raven puts it: “Even the smartest computers are still pretty dumb. We want to beat up the theory that we’re all going to die from robots. Experts and skeptics in the space of AI, experience design, and innovation will talk about the ‘useless’ things that are being made with AI as well as the genuine advancements in a conversation that will examine where the AI apocalypse is really at, and if it’s even ever going to happen.”

Perhaps the PR and marketing experts behind this talk have copped the number of AI cynics who will be in attendance this year and I'm simply falling into another one of their traps. I'll let you know if I leave wearing an AI – Still Terrible t-shirt and cap.

Cynicism is often mistaken for realism. In this case, it’s not. There’s no reason AI can’t become the game-changing tech many proselytise it to be. It has limitless potential at this point. But, like The Beatles, it’s not “bigger than Jesus” despite what anyone says.