IBM chief executive Arvind Krishna says AI payoffs need companies to rip up old workflows, not just sprinkle bots around.
Krishna said that the use of AI inside a company usually moves from individual workers to small teams, cross-functional groups and then the whole organisation.
As a company moves from one stage to another, the potential return on AI investment increases. Krishna said that it depends less on tech and more on fixing ancient processes and office politics.
“In the next year or two, the enterprise world will sort into two camps: companies where AI runs their business, and companies where AI is still a project,” Krishna said.
The difference between firms that make broad use of AI and those still playing with pilots will not come down simply to technology, but become their operating model, Krishna said.
Biggish Blue plans to use its Think conference in Boston on Tuesday to announce a clutch of products and capabilities.
These include a new version of watsonx Orchestrate, a secure multiagent control plane and IBM Bob, for securely building and deploying agents.
While model developers race to keep a nose ahead, Big Blue is pitching AI from another angle. It wants to help clients scale their AI efforts without the usual corporate soup.
Last month, IBM reported higher first-quarter revenue of $15.92 billion and higher profit, driven by growing adoption of artificial intelligence tools.
The numbers beat expectations, but the share price still fell, because investors can find a cloud in any spreadsheet.
Part of that reflects IBM being dragged into wider AI-driven worries about software.
Krishna pointed to IBM’s own human resources processes as an example of the operating model needed for enterprise-level scaling.
Before AI, Krishna said an employee seeking an employment verification letter for an apartment rental application faced a daft workflow.
The process required up to 18 human touchpoints, including a manager, an HR business specialist, back-office staff and several software systems. Now an employee can generate the letter by asking an internal bot called Ask HR.
The AI agent is integrated into IBM’s security network. It verifies the employee’s identity, pulls the required data from the HR system and sends the letter.
The employee only has to say how the letter should be delivered. According to Krishna, the 18 touchpoints have been cut to one.
Krishna said processes must be redesigned end-to-end if AI is to scale across the enterprise.
IBM’s Project Bob, as it was known internally before becoming a product, was built to manage the full software development life cycle.
That covers writing new code, patching old code, generating documentation, creating test cases and checking security compliance, Krishna said.
“I don’t begin with eliminating steps. I begin by asking how many touchpoints I can take out. And how can I make it much more nimble and faster and end-to-end? That’s the goal. Out of that comes the fact that you should eliminate steps,” Krishna said.
If people lose their jobs because of the shift, “I get to redeploy them, to do something else of more value,” Krishna said.
