I’ve been writing and asking readers about your experiences with digital innovation and disruption for some time now, and think it worthwhile for us to stay on top of such matters. Artificial intelligence (AI) has made substantial news this week, and so I’m bringing it to you in the form of this latest Weekend Poll.
US Congress hearings on AI
On June 26th, the US Congress held hearings on the future of artificial intelligence (AI). Greg Brockman, CTO and co-founder of OpenAI, was one of four witnesses to testify. OpenAI is a non-profit AI research company, which has a stated goal of “discovering and enacting the path to safe artificial general intelligence” – to ensure that AI can co-exist with humanity and, as its website states, to ensuring AI leads to a good outcome for humans.
If you’ve not hear of OpenAI before, you’ll have heard of one of Brockman’s fellow co-founders, Elon Musk. The organisation is sponsored by individuals – and by Amazon Web Services, Infosys, Microsoft, Open Philanthropy Project and YResearch (Y Combinator Research). OpenAI’s mission is “to build safe AGI, and ensure AGI’s benefits are as widely and evenly distributed as possible”.
The field of AI is moving faster than we realise
Brockman was in demand this week, and I watched an interview he gave to BNN Bloomberg. In the interview, Brockman said that the field of AI is moving very fast, a lot faster than people realise.
“We’re starting to be able to build faster and faster computers, at a rate that is a lot faster than anything we’ve seen – and so we don’t know, five years from now, what we will be capable of building.”
Brockman explained that OpenAI recently studied the growth of computation in the field, and “found that the amount of compute thrown at the largest AI systems has been growing at a rate of doubling every three and a half months for the past six years.”
“That’s a total growth of 300,000 times and we project the same thing for the next five years, which means that, to put that in perspective, if would be like if your phone battery went from having one day of battery life to having 800 years of battery life, and then five years from now and (having) 100 million years of battery life.”
AI has begun beating humans at esports
Now, I don’t follow esports, but the International DOTA 2 Championships are a big deal. Now in its eighth year, the 2018 prize pool is $16M US.
OpenAI has created AI bots that can compete against humans in eports gaming … and win. That’s just what OpenAI Five did this week, playing video game DOTA 2, right here in Canada. As Bill Gates put it, “That’s a big deal, because their victory required teamwork and collaboration – a huge milestone in advancing artificial intelligence.”
In his BNN Bloomberg interview, Brockman elaborated on how this would represent a milestone: “.. unlike board games like Chess or Go, these strategy games start to capture these complex aspects of the real world. You need to be able to strategize over this very continuous time. What we’ve been able to do is show that today’s algorithms, if you scale them up with massive amounts of computation, are capable of going far further than even the experts in the field thought.”
“Today, we’re able to build systems that are able to plan and reason over these long time horizons, that are able to operate in these very complex environments – something very different than what we’ve seen before. And so you start to think about potential applications …. that can be deployed in the real world, helping real people; that are able to navigate and operate in these very critical, very high responsibility areas.”
Systems that can be deployed in the real world
I know that many assistants don’t see AI impacting their current roles, and understand that. It’s worth noting, though, Brockman’s comments about potential AI applications that can operate in areas of high responsibility. When asked about the impacts of AI in the short and long terms, Brockman offered his take: people overestimate the immediate (short term) impacts of technologies, but underestimate the long term impacts.
How do people feel about robots and their jobs?
We have data out of the US, where Pew Research Centre asked American adults last year if they were worried that robots and computers could replace them in their jobs.
The results? 73% of those surveyed reported that they were either somewhat (48%) or very (25%) worried. Just shy of a quarter of respondents (23%) reported that they were “not too worried”, while another 4% said they were “not at all” worried. The results noted that “Americans think automation will likely disrupt a number of professions – but relatively few think their own jobs are at risk.”
All this leads to my latest Weekend Poll question, in which I’d like to compare your views with those reported by the Pew Research Centre.
Are you worried that robots and computers could replace you in your job?
Please take a couple of minutes to complete the poll below. As always, I look forward to hearing what you have to say and will publish results early next week.