The complicating factor for investors is that IBM
won’t say how much of its revenue is tied to AI.
Krishna says the number is impossible to
compute. “Our mainframes have AI circuits—so
is the mainframe business AI? I am going to do
storage backup in the future with AI—is storage
backup AI? Maintenance applications use AI. All
of cybersecurity is going to use AI. Before the
next five years are done, everything is going to
have AI fused inside it.”
What a weaselly response.
https://www.barrons.com/articles/ai-stocks-dividend-ibm-99e61b50
By: Eric J. Savitz
June 2, 2023 5:38 pm ET
One of the most alluring bets on artificial intelligence is hiding in plain sight. It’s a company everyone knows—a tech giant that has focused its business for years around the cloud and AI. It also has one of the highest dividend yields in the tech sector.
This AI surprise is... IBM .
There are reasons the market is skeptical about Big Blue’s position in AI, but the omission is shortsighted. Shares of chip maker Nvidia (ticker: NVDA) are up 172% this year, driven by the success of its chips used in generative artificial intelligence applications. Microsoft (MSFT) and Alphabet (GOOGL), the leading players in AI software, have added 40% apiece.
IBM (IBM)? It’s down 8%.
That surprising decline comes even as IBM has been remaking itself over the last three years under the steady hand of CEO Arvind Krishna. IBM has arguably deeper AI knowledge than almost any company, and yet the stock trades at less than two times forward sales, versus Nvidia at nearly 20 times. IBM also has a dividend yield above 5%.
The company has been working on AI applications for at least four decades. In 1997—two years before Nvidia came public—an IBM supercomputer called Deep Blue defeated world chess champion Garry Kasparov in a six-game match. In 2011, IBM’s Watson supercomputer famously defeated human champions on Jeopardy. Among Watson’s adversaries was the legendary Ken Jennings, now the show’s host. “I, for one, welcome our new computer overlords,” Jennings wrote below one of his Final Jeopardy answers.
To be sure, IBM has made missteps. After the Jeopardy stunt, it made a big push to use Watson in healthcare, for dr-g discovery and other applications. That never quite worked, and IBM sold the Watson Health business for a reported $1 billion in 2022. Some people seem to have interpreted the Watson Health sale as IBM giving up on AI, but that’s far from the case.
This past week, I had a long conversation with Krishna about the company’s approach to the AI business. He had a lot to say.
For one thing, IBM recently launched an all-new version of Watson called Watson X. The new offering has three parts: Watson.ai works with customers to create new models, or data sets. Watson.data acts as a data store, putting the company in competition with Snowflake (SNOW), among others. And Watson.governance monitors AI models to make sure they are accurate and accountable, not filled with false or offensive information.
What IBM isn’t going to do is create the next ChatGPT. “Public models are incredibly powerful,” Krishna said. “What Google, Facebook, Microsoft are doing absolutely fits that mode. They are building very, very large models that serve everybody.”
But Krishna thinks the public-facing AI applications are just a small portion of the opportunity. “It’s like an iceberg,” he says, with chatbots such as Microsoft Bing and Google Bard above the waterline. “There are more use cases that are not going to benefit from a large public model.”
IBM’s strategy is to help customers create AI applications of their own, to squeeze more value out of their data. In some cases, the company blends open source models with proprietary data. For some customers, IBM is building private models specifically for their data.
While IBM has no plans to build a general large model like ChatGPT, Krishna says it is building out a family of domain specific data sets. For instance, it’s creating a chemistry model, based on public domain information.
“If I take one of my chemical industry partners—let’s say Dow, Mitsui Chemicals, or BASF—they have proprietary data on how to manufacture chemicals,” he says. “Would any of them put their proprietary data into a public model? Of course not. But they would love to extend the chemistry model to include their own data...they could deliver speedier answers to client and internal questions, and come up with new formulations. It will be an accelerant to their business model.”
In one case, Krishna says, IBM is working with a bank to come up with better compliance and audit data for its own people, in part to show regulators that it has the right controls in place.
IBM has built 20 of these domain models—beyond chemistry and banks, the list includes one to write code, another to make IT operations more efficient. There’s a geospatial model IBM is adapting to combine with NASA climate data, to improve weather modeling. The company—which owns the open source software giant Red Hat—isn’t shy about using open-source training models. In particular, IBM has partnered with Hugging Face, an AI start-up with a library of 130,000 models.
The complicating factor for investors is that IBM won’t say how much of its revenue is tied to AI. Krishna says the number is impossible to compute. “Our mainframes have AI circuits—so is the mainframe business AI? I am going to do storage backup in the future with AI—is storage backup AI? Maintenance applications use AI. All of cybersecurity is going to use AI. Before the next five years are done, everything is going to have AI fused inside it.”
Krishna does cite a PwC estimate showing that AI could generate $16 trillion in value for the economy by 2030. And Krishna has a longstanding view that IBM over time should grow revenue in the mid single digits.
What the CEO won’t directly say is that AI opens new opportunities for IBM and could be additive to sales. “It’s a potential accelerant,” he says.
Investors should read between the lines.