Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

IBM’s Watson Health is sold off in parts

statnews.com

Post reply on HN
Print view View original

691–700 of 711 posts

Re: IBM’s Watson Health is sold off in parts

#691

Post by listenallyall »

politician wrote: post #690
listenallyall wrote: post #172

Earlier quoted context omitted.

So all you're saying is that cryptocurrencies can be deployed to facilitate illegal behavior. We've known that for quite a while.

To interpret a tool like prediction markets as you have, is it fair to say that you strongly believe in the divine right of kings?

I never said that prediction markets should be illegal -- I'm just acknowledging what you stated, that in some jurisdictions, they are. That's a fact. Running a cryptocurrency-based prediction market doesn't make it less illegal.

Your argument isn't with me, it's with whoever decided a prediction market should be illegal. Go change that person's mind.

Re: IBM’s Watson Health is sold off in parts

#692

Post by Phillipharryt »

School-Cotton wrote: post #684
Phillipharryt wrote: post #679

Earlier quoted context omitted.

That is a form of machine learning. It uses an algorithm with an enormous ability to look ahead, and selects the moves according to a heuristic that have been shown previously to be most likely to create the desired end state. Chess is not a solved game from the starting state, so it has to make assumptions based on the data it has. This is machine learning. Please don't make such definite statements. You even say it…

> It uses heuristics to assign values to different board states, and in the case of Deep Blue these values were created through previous game analysis. Unless I’m misinformed, this part isn’t true. The heuristic was hardcoded with the help of human experts.

From the paper Deep Blue by Murray Campbell et al (people who worked on Deep Blue) "The initialization of the feature values is done by the “evaluation function generator”, a sub-program which was run on the master node of SP system." Which would suggest it generated the heuristic itself. The features may have been hardcoded, but assigning values to them wasn't. In addition to this, feature values could be static or dynamic, meaning it would update dynamic ones depending on the board state. It not only generated the heuristic values (feature values) it could modify them to reflect their relative change in impact throughout the game.

Re: IBM’s Watson Health is sold off in parts

#693

Post by Traster »

bambataa wrote: post #649
doctor_eval wrote: post #566

Earlier quoted context omitted.

If I remember correctly, John Ralston Saul made the point [0] that there is little observed difference in terms of the efficiency of decision making between large government and large corporations. My personal experience is that large organisations and government are barely distinguishable. This is often excused in the name of "risk mitigation" - but in my experience it's really just that there are more snouts in the…

This is been my anecdotal experience. In the UK, I note that politicians extolling the efficiency of the private sector rarely have any substantial experience in big companies.

They have plenty of experience being wined and dined by executives of big companies though.

Re: IBM’s Watson Health is sold off in parts

#694

Post by mattkrause »

listenallyall wrote: post #608
matheusmoreira wrote: post #605

Earlier quoted context omitted.

There's more material available, not just scientific journal articles. The medical specialty associations periodically release detailed guidelines covering new criteria, methods, treatments. For example: https://www.ahajournals.org/doi/pdf/10.1161/HYPERTENSIONAHA.... There are courses available if you wish to be taught. There are events where people present new developments. There are paid sites like UpToDate. Even t…

Did you check the Disclosures section of your article? Are there any drug companies NOT included there? Like it or not, private companies fund the vast majority of medical research in the U.S., and any doctor who wants to stay up to date cannot (and should not) ignore it all. The authors have no conflict of interest to declare, but declare lecture honoraria or consulting fees as follows: T.U., Bayer, Boehringer Ingel…

The NIH funds a huge fraction vast of biomedical research in the US. Pharma does important work too, but it’s largely concentrated in the very last stages of getting a product to market.

These COI disclosures also strike me as hard to interpret. It’s certainly possible that some of these people are deeply invested in a company and are pushing its particular therapy hard to buy a new boat or something. However, I’d bet many of them are $250 to participate in a focus group, or free conference registration to be in a panel. It’s important to know who’s buttering the authors’ bread, but it’d be helpful to know how much it’s being buttered too.

Re: IBM’s Watson Health is sold off in parts

#695

Post by kbelder »

Firmwarrior wrote: post #294
colinmhayes wrote: post #221

Earlier quoted context omitted.

Of course AI could replace doctors. Some people will always prefer the bedside manner a person provides, but many others won't care, and AI will eventually do a good job at imitating a good doctor's bedside manner.

The only way AI is going to replace doctors is if medical technology advances to the point where you can repair a human as well as you can repair a machine (or someone invents AGI, I guess) There are just too many unknowns and fiddly things going on with human bodies edit: Haha, I guess Tex's sibling comment makes a good point though..

I don't think we've got to the point where AI can repair a car. Humans are more complicated by a decent factor.

Re: IBM’s Watson Health is sold off in parts

#696

Post by mattkrause »

krageon wrote: post #661
matheusmoreira wrote: post #614

Earlier quoted context omitted.

I did see that. I don't see that as a reason to doubt the recommendations of this particular article. I thought we were talking about direct marketing by pharmaceutical company representatives by the way. It is of course impossible to separate modern medicine from the pharmaceutical industry since many therapies depend directly on them. It's still the doctor's job to figure out which medicines are actually good and w…

> I don't see that as a reason to doubt the recommendations of this particular article. No conflict of interest, except they're directly paid by drug companies. That's the sort of thing you'd expect to see in a banana republic, not established medicine. It should call the integrity of the entire system into question, frankly.

What’s a better solution though?

Do you really want to ban all interactions between researchers/clinicians and drug companies?

Re: IBM’s Watson Health is sold off in parts

#697

Post by listenallyall »

mattkrause wrote: post #694
listenallyall wrote: post #608

Earlier quoted context omitted.

Did you check the Disclosures section of your article? Are there any drug companies NOT included there? Like it or not, private companies fund the vast majority of medical research in the U.S., and any doctor who wants to stay up to date cannot (and should not) ignore it all. The authors have no conflict of interest to declare, but declare lecture honoraria or consulting fees as follows: T.U., Bayer, Boehringer Ingel…

The NIH funds a huge fraction vast of biomedical research in the US. Pharma does important work too, but it’s largely concentrated in the very last stages of getting a product to market. These COI disclosures also strike me as hard to interpret. It’s certainly possible that some of these people are deeply invested in a company and are pushing its particular therapy hard to buy a new boat or something. However, I’d be…

> The NIH funds a huge fraction vast of biomedical research in the US.

This document https://www.researchamerica.org/sites/default/files/Policy_A... states that in 2017 private industry spent $121 billion on Medical & Health R&D Expenditures in 2017 compared to $39 for the federal govt ($32 of which is NIH). $121 billion in a single year -- I honestly assumed the number was in millions till I re-read it.

This article https://www.baltimoresun.com/health/bs-hs-trial-funding-2015... says "Industry funds six times more clinical trials than feds"

You are probably right that the drug companies are (not surprisingly) focused specifically on drugs while NIH research is more general and widespread. But that is still a very large difference.

Re: IBM’s Watson Health is sold off in parts

#698

Post by nextaccountic »

listenallyall wrote: post #568
nextaccountic wrote: post #545

Earlier quoted context omitted.

> How else are doctors, especially those who are 10, 20 years removed from their residency, supposed to learn about new treatments and medicines? With some other mechanism that doesn't have major conflicts of interest? Like, I don't know, attending to actual courses, like any other profession that requires continuing training.

Who is teaching the courses? Where do they get their info? Given the very limited oversight of "continuing education", it's trivial for Merck or Pfizer to ensure their advocate teaches the class or records the online seminar -- with zero requirement that they disclose that fact. At least when a drug rep visits the office and buys lunch, you know the source.

At this point we need public policy to stop such bad actors. Or you mean we should give up and do nothing? But why?

Re: IBM’s Watson Health is sold off in parts

#699

Post by chaostheory »

rprospero wrote: post #665
chaostheory wrote: post #607

Earlier quoted context omitted.

> An "AI chatbot" is far inferior to a real user interface. That's because you're lucky. You have good enough sight and you can use your hands. Unfortunately, that's not the case for millions of people especially since our populations are aging more.

I'm curious about your response here, as I'm certain that I'm missing something. I have cerebral palsy, which comes with issues using my hands. I've also had noticeable degradation with age of my eye-sight over the past couple of years. That said, I'm not sure how a chatbot helps people with these issues. Typing multiple sentences into the chatbot interface is certainly more tiring and error prone that click a few bu…

Alexa, Siri, and Google Home are essentially AI chat bots using microphones. The former two can barely process anything you ask but you can still get a lot done with them.

Re: IBM’s Watson Health is sold off in parts

#700

Post by Shocka1 »

formeribmer wrote: post #479

I worked at IBM Watson as one of the early engineers when they first started commercializing the product. It was a fucking joke - Ginni Rometty would go up on stage and said that Watson can help diagnose cancer from CT scans and we would just look at each other and be like "Dude, Watson is just a glorified Lucene index, wtf is she talking about." They started selling Watson as the end-all for everything from cancer d…

I worked for IBM (non-Watson) not too many years ago and this cracks me up because I heard the same from Watson employees I spoke with. And it wasn't just Watson that Ginni talked fluff about. I swear every word that came out of her mouth was a combination of fluff and technical buzz words. We all used to sit around a workstation when she would give her monthly talks and ask one another what she was talking about, because none of us had a clue. IBM seems to encourage that fluffy, buzz word exec talk, and to be honest if I hear someone talk like this now my mind turns off.

One of my favorite past-times is talking with loyal IBM employees and seeing if they can tell me what Watson is, does, or how to use it. Ginni is laughing all the way to the bank at this point so I guess the joke's on us.

Post reply on HN