The arrival of advanced artificial intelligence (AI) in cloud networks, and the potential consequences of its adoption for the human workforce, is starting to make a lot of people very nervous – including me. That’s not surprising, given announcements like the one from BT last month, where it said it plans to use AI as part of a strategy to cut 55,000 workers by the end of the decade.
Oof.
But how worried should we really be?
That’s the question I posed to Matthew Jones, chief architect of Red Hat’s Ansible Automation Platform, last week at its event in Boston. And you know what? I actually felt a lot better about humanity’s prospects after our conversation.
Jones told me he originally comes from Mississippi, which means I had the opening bars of Elvis’ “Suspicious Minds” stuck in my head for the whole interview. More importantly, he was one of the original coders on Ansible and has been responsible for its development throughout its journey to last week’s big co-announcement between Red Hat and IBM adding generative AI to Ansible.
“People who work at Red Hat come from a different place to some other organizations. We’re big on humanity,” said Jones, who pointed out that Ansible is named after the faster than light communication technology in Ursula Le Guin’s science-fiction novels. These, of course, are known for their humanistic themes (and giant flying cats, which is awesome, obviously).
It’s clear from talking to Jones that Red Hat’s approach to AI stems from its ‘sharing is caring’ open-source culture, where people (emphasis on ‘people’) have tended to played more nicely than other parts of the communications industry.
“I don’t see the humans being moved out. We’re helping to deal with human problems, and we want them in the loop,” Jones said. “We see AI as a tool to help organizations do more with less, and to eliminate the repetitive tasks that people don’t want to do anyway.”
Pay to play
Red Hat is going through a major transformation itself right now, as it shifts to asking users to pay for products like Red Hat Enterprise Linux (RHEL) and Ansible, rather than getting its tech through free community distros.
I think it will succeed here, for a few reasons:
First, Red Hat had a particularly strong show in Boston, making a series of significant announcements, including several with major enterprise customers. Its partnership with IBM also looks like a powerful and productive mind meld.
Then there's its Ansible automation architecture, which divides the automation engine and the automation functions. This structure makes a lot of sense for vertical industries, since it allows the industry experts within each vertical to create the specialized automation they need rather than having a cloud supplier try to cobble it together for them. Once created, these specialized tools can then be hooked back into Ansible through APIs.
But I also keep coming back to Jones’ comments about the importance of keeping people in the cloud decision-making loop. If I was running cloud infrastructure for a major enterprise or telco, I’d want to buy my technology from a company that was thinking big picture thoughts about the potential outcomes of AI and automation.
Do I think other cloud companies, such as Oracle, are doing the same? No, I don’t.
Larry Ellison’s reputation suggests he doesn’t spend a lot of time thinking about redundancy risks to people lower down the org chart. And Meta is already run by a robot, so putting the survival of human cloud ops teams ahead of fiscal performance is unlike to compute for it.
So far, Red Hat is getting AI and automation right. Kudos to them, and Matthew Jones.