Close

IBM Invests $1 Billion to Grow Watson Supercomputer's Struggling Business

IBM is investing $1 billion in its Watson supercomputer in hopes of building a $10 billion business within a decade.

2 min read
IBM Invests $1 Billion to Grow Watson Supercomputer's Struggling Business
IBM

IBM faces a puzzle that might give even the world's greatest detective pause: How to grow its "Jeopardy"-conquering Watson supercomputer into a $10-billion business within a decade. The company hopes to boost Watson's profit-making possibilities with a $100 million venture-capital fund aimed at spurring new apps based on the computer's problem-solving capabilities.

Watson achieved fame by beating human "Jeopardy" champions in several 2011 matches that showcased the computer software's learning capabilities. But an Oct. 2013 conference-call transcript (reviewed by The Wall Street Journal) reveals that Watson has struggled to translate its "Jeopardy" sleuthing into real-life problem-solving.

The IBM Watson business unit has generated less than $100 million so far. But the slow start has not stopped IBM from aggressively investing in Watson as a cornerstone of its future business. The company recently announced it is investing more than $1 billion in its Watson business unit and setting up a $100-million venture capital fund to encourage apps built on Watson's technology.

During the Oct. 2013 conference call, IBM CEO Virginia Rometty set forth the target of having Watson generate $10 billion in annual revenue within 10 years. She based her goal on the projection that Watson would earn $1 billion in revenue per year by 2018.

Watson's first customers hope to leverage the supercomputer's learning capabilities. In one case, Watson is working on recommending potential cancer treatments for specific medical cases based on a confidence percentage. The supercomputer learns from its mistakes when human physicians correct its errors during training.

But translating individual customers' business technicalities into usable software for Watson to process has been rough sledding for IBM engineers. That has led to delays in one of Watson's biggest client projects—an effort, involving the University of Texas M.D. Anderson Cancer Center, to make a version of Watson capable of recommending leukemia treatments by sifting through medical journals and books.

A similar healthcare application is in the works in conjunction with the Memorial-Sloan Kettering Cancer Center in New York City, where Watson would also serve as an adviser on cancer diagnoses and treatments. Watson has also found work with Wellpoint, the U.S. largest health insurer.

Turning Watson's "Jeopardy" smarts into huge business profits may seem anything but elementary for IBM at this point. But to the company's credit, it has continued to back Watson despite the early difficulties and has enlisted AI researchers to continue improving the supercomputer's capabilities as it reaches out to more potential customers.

Photo: IBM

The Conversation (0)

The Future of Deep Learning Is Photonic

Computing with light could slash the energy needs of neural networks

10 min read

This computer rendering depicts the pattern on a photonic chip that the author and his colleagues have devised for performing neural-network calculations using light.

Alexander Sludds
DarkBlue1

Think of the many tasks to which computers are being applied that in the not-so-distant past required human intuition. Computers routinely identify objects in images, transcribe speech, translate between languages, diagnose medical conditions, play complex games, and drive cars.

The technique that has empowered these stunning developments is called deep learning, a term that refers to mathematical models known as artificial neural networks. Deep learning is a subfield of machine learning, a branch of computer science based on fitting complex models to data.

Keep Reading ↓ Show less