IBM’s Watson Health is sold off in parts

IBM identified back in the 70s their core asset and were very explicit about what it was. "No one gets fired for buying IBM." Their key strategy since then has been to monetize this asset in a variety of creative, and mainly very effective ways.

Thinkpads were a great example of this. Laptops which promised decent quality and support for a high price. When the laptop market was demystified and commodified, IBM correctly got out of it - for a decent price.

If some random start-up, or even Google, had built Watson, it would have correctly been seen as a gimmick. Instead it sold literally billions of software consulting to people who thought they needed AI but actually just needed a search box with dynamic autosuggestion. Would you rather get some junior guy to hack something together using open source tools, or would you rather pay IBM 50 times as much? If you chose the former, you're simply not in the target market.

The hybrid cloud is exactly the same game - as is made quite clear in that ad, it's pitched at middle management who don't want to look like chumps for ignoring the cloud, but don't want to fuck up by moving to it.

Reputation is a difficult asset to monetize - effectively you make money from it by degrading and then destroying it. After all, if you carry living up to your good reputation, you're not extracting any advantage from it. IBM can't sell their reputation or their name to the highest bidder. All they can do is keep trawling for business lines where it gives them a comparative advantage.

It's easy to see this as unscrupulous - but their customers genuinely do get a benefit from the confidence they have in IBM.

I think the person was suggesting using the results of the AI to inform the doctors, not replacing them. Which is something I would like as well

Here is a submarine Wired article on what Watson promoters expected it to do -- which supports what the parent commenter thought, the focus being on "bottom-line": https://www.wired.com/insights/2015/02/what-can-watson-do-fo...

Here is insight into how non-profit and community health programs are funded: https://www.parklandhospital.com/financial-summary

Here is a section of Healthcare new dedicated to AI/ML implementations and other related events: https://www.healthcareitnews.com/topics/artificial-intellige...

Here is CMS's directive for hospital price transparency: https://www.cms.gov/hospital-price-transparency

The parent comment assumed that everything in healthcare is profit-motive driven. However, there are large portions of the healthcare industry that are non-profit, that are transparent in their funding and their costs, and that are looking to implement AI to improve healthcare outcomes. Parkland Health and Hospital System, Harris Health System, University Health System are some that I am more familiar with that run with this (PHHS recently achieved HIMMS Level 7 certification, for example). These are social safety-net hospitals and healthcare systems -- they care for everyone regardless of ability to pay. They focus not only on emergency and inpatient care, but also ambulatory care, primary care, and even fund (at arm's length) community (non-profit) medicaid insurers.

On a more subjective side, I've seen a lot of folks out to make a buck, but the non-profit healthcare side has been much more focused on patient outcomes.

Many studies comparing NP and physician outcomes will have the NPs under supervision by physicians, which is ideally how they would be used, but in practice the true supervision level varies widely. I wouldn't see an NP for my care personally, and I doubt there are many physicians who would. The wait time to see primary care physicians is typically less than a week in most places and would be worth it. If you're experiencing something you feel is too serious to wait a week I would visit the ER (and make sure to ask to be seen by the physician also). It's your health. Personally I would only trust mine to the people who are the experts in their subjects, and not those who have less training and can switch between specialties without any additional training.

I don't have anything against NPs when the supervision is close, but more and more doctors are put into positions where they are acting as liability sponges for de-facto independent NPs/PAs.

Here are a few studies - (CRNA) We found an increased risk of adverse disposition in cases where the anesthesia provider was a nonanesthesiology professional. https://www.ncbi.nlm.nih.gov/pubmed/22305625

Comparing urgent care visits between MD/DOs and Midlevels. Doctors saw more complicated patients, addressed more complaints and deprescribed more. https://link.springer.com/article/10.1007/s11606-021-06669-w

NPs/PAs practicing in states with independent prescription authority were > 20 times more likely to overprescribe opioids than NPs/PAs in prescription-restricted states. https://pubmed.ncbi.nlm.nih.gov/32333312/

Both 30-day mortality rate and mortality rate after complications (failure-to-rescue) were lower when anesthesiologists directed anesthesia care. https://pubmed.ncbi.nlm.nih.gov/10861159/

Compared with dermatologists, PAs performed more skin biopsies per case of skin cancer diagnosed and diagnosed fewer melanomas in situ, suggesting that the diagnostic accuracy of PAs may be lower than that of dermatologists. https://www.ncbi.nlm.nih.gov/pubmed/29710082

Advanced practice clinicians are associated with more imaging services than PCPs for similar patients during E&M office visits. https://jamanetwork.com/journals/jamainternalmedicine/fullar...

Nonphysician clinicians were more likely to prescribe antibiotics than practicing physicians in outpatient settings, and resident physicians were less likely to prescribe antibiotics. https://www.ncbi.nlm.nih.gov/pubmed/15922696

The quality of referrals to an academic medical center was higher for physicians than for NPs and PAs regarding the clarity of the referral question, understanding of pathophysiology, and adequate prereferral evaluation and documentation. https://www.mayoclinicproceedings.org/article/S0025-6196(13)...

Resident teams are economically more efficient than MLP teams and have higher patient satisfaction. https://www.ncbi.nlm.nih.gov/m/pubmed/26217425/

Those are problems in theory, but I don’t think most people are concerned or affected by them. I’m certainly not.

I never need to transfer money quickly between accounts — it’s never once been a problem to wait a couple days. And sure there are real economic problems with printing money, but again, the government does a decent enough job at keeping the dollar stable that it doesn’t affect me.

And there’s the fact that the US gov has tons of power to maintain the validity of the fiat dollar through legislation, and as a backup they have the use of force through police and jail (in the case of tax evasion, or avoiding the laws). Then there are international alliances, and there’s the largest military in the world also with a strong interest in maintaining the dollar’s value.

So I’m not worried about the value of the US dollar in the long term — at least I certainly trust it more than a purely technical solution with none of the US Gov benefits.

Faster: I don’t have any problem with speed of USD transactions. In fact, most transactions are faster than crypto via credit cards or cash.

Cheaper: there are $0 transaction fees for cash, and low fees for credit.

Secure: US laws do a decent enough job

Fraud: crypto exchanges get hacked and there is often no recourse — if my credit card is stolen, there are laws that protect me

Limited supply: by definition, that makes the currency deflationary, which is horrible for a growing economy. And it’s obvious in bitcoin. Nobody spends money today if it’ll be worth more tomorrow — that’s why everyone just buys and holds bitcoin as an investment, not uses it as a currency