If AI is supposed to be all about trust why has IBM Watson seemingly been programmed to generate hype & operate like a dodgy accountant?


“A University of Texas System audit has found irregularities involving more than $40 million paid for outside goods and services as part of MD Anderson Cancer Center’s now stalled effort to enlist IBM supercomputer Watson in the battle against cancer…   …The audit, posted at the system’s website, reported the cancer hospital didn’t follow established purchase rules for the project, which executives had ballyhooed would tap Watson’s encyclopedic memory and lightning information-processing speed to tailor optimal patient treatment…   …The irregularities included service agreements that didn’t undergo a competitive process, fees consistently set just below the amount that would have required board approval…  …The project, which has incurred costs of $62 million, was placed on hold prior to the initiation of the audit…   …it’s unclear if any of the project’s costs contributed to operating losses at the renowned cancer hospital that have exceeded $430 million the past 16 months and that in January led to the layoff of 778 employees. More than half of the project’s funding came from gifts donated or pledged specifically for the purpose…   …The project — known as the Expert Oncology Adviser, powered by Watson — was unveiled by MD Anderson more than three years ago. Officials said then that doctors and staffers had spent the last several months feeding Watson unique case histories about more than 1 million of its patients, past and present, as well as disease information and research about treatment options. They said the knowledge would guide the cancer hospital — and ultimately worldwide — care…   …Now, not only is the system not in clinical use, it has not been piloted outside of MD Anderson, according to the audit. It adds that the agreement currently in effect states that the system “is not ready for human investigational or clinical use, and its use in the treatment of patients is prohibited” except as needed to test and evaluate the system. In addition, the project’s drug protocols and clinical trial data are now outdated and must be updated before the project can be piloted again within MD Anderson and with a network partner. The $62 million paid through Aug. 31, 2016 reflects external entities only, not internal resources, according to the audit”

Touted IBM supercomputer project at MD Anderson on hold after audit finds spending issues by Todd Ackerman in the Houston Chronicle.

Guruduth Banavar, Chief Science Officer & Cognitive Computing Vice President at IBM Research, is very clear in the AI Ethics paper they’ve published on Learning to trust artificial intelligence systems (“To reap the societal benefits of AI systems, we will first need to trust it. The right level of trust will be earned through repeated experience, in the same way we learn to trust that an ATM will register a deposit, or that an automobile will stop when the brake is applied. Put simply, we trust things that behave as we expect them to“) so it baffles me that after all the hype of IBMWatson it seems to have materially done very little/anything beyond helping generate gift pledges (probably from tech firms and Digital Health investors associated with trying to make the Digital Health Clinic powered by IBM Watson experience happen).


To end on a positive at least those involved in this project feeding the data in probably at least know what Big Data actually is (unlike contemporaries in the NHS etc).


*** Update 13 November 2017 ***

“In reality, Watson for Oncology is a “mechanical turk” — a human-driven engine masquerading as an artificial intelligence. The way it actually works is by convening a small panel of cancer experts from Memorial Sloan Kettering Hospital, who come up with recommendations for specific patient profiles. These recommendations represent the best guesses of these experts, supported by medical literature and personal experience. 


IBM has never allowed an independent study of Watson for Oncology. No followup is done to evaluate whether its recommendations help patients.

There are several problems with this approach. First, there is the deceptive marketing of Watson for Oncology to doctors and patients, who believe they are getting a global, data-driven, empirical recommendation, as opposed to the subjective judgment of a small panel of experts”

About David Doherty

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s