Cognitive computing, or machine learning, as exemplified at IBM by Watson, is technology that learns as it goes, specifically by ingesting and crunching lots of data. Computer systems like these learn on their own, without human intervention. IBM is building businesses in healthcare and other areas around Watson.

IBM (IBM) chief executive Ginni Rometty said the company sees a huge opportunity—$2 trillion worth—in this market over the next decade. And that’s above and beyond the $3 trillion opportunity in more traditional IT gear like servers, software, storage boxes and the like, said executives. Rometty stressed, more than once, that IBM is now a “cognitive computing and cloud company.”

Get Data Sheet, Fortune’s technology newsletter.

These are huge market numbers, but as one analyst pointed out during the Q&A session, projections don’t necessarily translate to reality. IBM, he noted, also predicted a huge market in business process outsourcing a few years back, and that market didn’t meet expectations.

But the company’s brain trust was resolute about its contention during a question and answer session.

Asked where that $2 trillion would come from, Rometty characterized it as mostly a new market. “It’s not stealing from anything, but creating new value,” she said.

John Kelly, IBM’s senior vice president of cognitive solutions and research said most of that $2 trillion will come from eliminating waste in supply chain, research and development, and other processes across industries.

And, in healthcare, a market that’s becoming a big priority for IBM as evidenced by a $4 billion string of acquisitions, there is a lot of waste from missed diagnoses, lack of drug efficacy, and staffing. “We think $200 billion of that is addressable by cognitive solutions,” Kelly said.

For more on IBM’s healthcare acquisitions, watch this video.

But cutting waste is just the easy stuff, in his view. “I think there may be something larger around that, real opportunity around making really good optimized decisions,” he noted. If companies can use cognitive computing to build new services, that’s additive business.

Rometty added that cognitive solutions around weather prediction could save insurance companies big money. Towards that end, earlier this quarter IBM completed its buy of the Weather Co. and its data trove.

Another huge opportunity for cognitive computing is preventive maintenance in manufacturing and heavy industry. The idea there is that industrial machinery equipped with connected sensors can provide continuous feedback on performance, so upkeep can be scheduled before there’s an actual breakdown.

IBMers also claimed blockchain technology, the distributed transaction engine behind Bitcoin, as another big potential market. Bitcoin itself represents probably 1% of blockchains potential usage, said Arvind Krishna, senior vice president and director of research.

Is IBM’s Transformation for Real?

Other applications could include tracking property records, which is a huge concern in developing markets where proving land ownership is difficult, and leads to fraud and corruption. “Having a blockchain that is immutable, i.e. can’t be changed, to prove who owns what is extremely important,” he said.

Wider application of blockchain, which IBM now offers as a service from its Bluemix cloud-based software development platform, can speed up transactions and suck out the cost of doing business, he said.

So IBM’s top execs set up some mighty big expectations on market size and opportunities. But huge questions loom. Chief among them: Even if that multi-trillion market does come to fruition, just how much of that opportunity will accrue to IBM rather than to Microsoft (MSFT), Amazon (AMZN), Google (GOOG) or others?