Hundreds of AI Tools Were Built to Catch Covid. None of Them Helped (technologyreview.com) 108
At the start of the pandemic, remembers MIT Technology Review's senior editor for AI, the community "rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines — in theory.
"In the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful." That's the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK's national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid.
This echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Laure Wynants, an epidemiologist at Maastricht University in the Netherlands who studies predictive tools, is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing. "It's shocking," says Wynants. "I went into it with some worries, but this exceeded my fears."
Wynants's study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use. "This pandemic was a big test for AI and medicine," says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. "It would have gone a long way to getting the public on our side," he says. "But I don't think we passed that test...."
If there's an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. "The pandemic has put problems in the spotlight that we've been dragging along for some time," says Wynants.
The article suggests researchers collaborate on creating high-quality (and shared) data sets — possibly by creating a common data standard — and also disclose their ultimate models and training protocols for review and extension. "In a sense, this is an old problem with research. Academic researchers have few career incentives to share work or validate existing results.
"To address this issue, the World Health Organization is considering an emergency data-sharing contract that would kick in during international health crises."
"In the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful." That's the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK's national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid.
This echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Laure Wynants, an epidemiologist at Maastricht University in the Netherlands who studies predictive tools, is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing. "It's shocking," says Wynants. "I went into it with some worries, but this exceeded my fears."
Wynants's study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use. "This pandemic was a big test for AI and medicine," says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. "It would have gone a long way to getting the public on our side," he says. "But I don't think we passed that test...."
If there's an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. "The pandemic has put problems in the spotlight that we've been dragging along for some time," says Wynants.
The article suggests researchers collaborate on creating high-quality (and shared) data sets — possibly by creating a common data standard — and also disclose their ultimate models and training protocols for review and extension. "In a sense, this is an old problem with research. Academic researchers have few career incentives to share work or validate existing results.
"To address this issue, the World Health Organization is considering an emergency data-sharing contract that would kick in during international health crises."
Not deep learning (Score:2, Informative)
There aren't enough patients sick yet of covid for deep learning. The problem with deep learning in the medical field is the small size (relatively) of the data sets.
Re: (Score:2, Insightful)
Re: (Score:1)
Re: (Score:2)
Re: (Score:2)
Then we'll call it Dojo [youtu.be] then.
This is not insightful. (Score:2)
The intelligent human comes about as the result of processing trillions of data points.
Re: (Score:2)
Unlike intelligent doctors (Score:2)
The idea that you need millions of points of input data in order to produce a result is the opposite of intelligence.
Yeah, whereas med students are intelligent because they can immediately give out perfect diagnostics immediately after they walk out of the lecture~
You know there is this metaphor about needing to practice something for 10'000 hours of anything before being good at it ?
The whole idea behind deep-learning methods is to have an artificial neural construct going through the 10'000 hours in simulation.
The parent is pointing out that this is going to be hard when all you have is the equivalent of a metaphorical
Re: Not deep learning (Score:2)
One thing is certain, this kind of trite old comment is not intelligent.
You can't point to anything that our brains definitively does, that a learning algorithm definitively doesn't. Few-shot learning through pretraining is one of the big themes in research the last couple of years. We can make systems that learn from few examples, like humans do. Of course then we get a bunch of biases from the pretraining as well, and start making more human-like mistakes as well. There's no free lunch in learning, whethe
Any thing I can do you can't do better (Score:2)
You can't point to anything that our brains definitively does, that a learning algorithm definitively doesn't.
According to this article: diagnose patients with COVID-19 and suggest treatment.
Re: (Score:1)
You seem to demand that our brains should be judged based on what the best and most specialized of us can do, in the best circumstances, but learning algorithms should be judged based on what exactly this one can do within exactly these constraints.
"If you speak to him of a machine for peeling a potato, he will pronounce it impossible. If you peel a potato with it before his eyes, he will declare it useless, because it will not slice a pineapple."
Re: (Score:2)
Re: (Score:2)
No you don't.
What you do need, for any reliable system including one involving humans, is good quality data, collected according to an organized protocol. Pretty much the opposite of the giant datasets people gather off the internet, and definitely not the chaos that came in through most of the pandemic.
Re: (Score:2)
For machine learning, you need good quality data. For deep learning, you need a lot of data.
If the dataset is small enough, machine learning isn't appropriate. You can use something like linear programming and get a precise answer.
Re: Not deep learning (Score:2)
Re: (Score:2)
Deep learning neural networks are, at their heart, regression models. You don't need a lot of data. You need a lot of data for *any* model that is very complicated. Deep learning models can easily be made arbitrarily complicated, therefore people who don't know how they work arrive at the conclusion that you need a lot of data.
There is a great deal of good work using deep learning on very small datasets.
Re: (Score:2)
If you don't have much data, then it's not deep learning, it's just learning.
Re: (Score:2)
Spoken like someone who doesn't have the slightest idea what the hell they're talking about.
Congrats. You are drowning in available information and successfully refuse to partake.
Re: (Score:3)
What are you talking about, there's probably a billion of people who've been sick with covid by this point.
AI is not AI. Duh. (Score:2)
AI is not really AI. Duh.
AI does not work. Duh.
Re: (Score:2)
There aren't enough patients sick yet of covid for deep learning. The problem with deep learning in the medical field is the small size (relatively) of the data sets.
There are plenty of patients... but there isn't enough data. Because health data is private.
A hospital can use data from its own patients (because it's allowed to use tools to aid in diagnosis of its own patients). But it can't distribute or publish the data, at least, not at the individual patient level needed to be useful to AI diagnosis; and it can't get data at that individual patient level from other hospitals, without each individual patient consenting to their data being transferred.
Re: (Score:2)
But it can't distribute or publish the data, at least, not at the individual patient level
Not even if it is anonymized? I don't think there are any
privacy laws restricting information on "Patient 7344901"
Re: (Score:2)
Tell it to the judge.
The last sentence is terrifying. (Score:3)
A bunch of bureaucrats want to create a new contract that forces folks under existing contracts to do what with data?
Re: (Score:2)
You get who you vote for.
No, I don't agree with that. We got who the vote counters selected, not necessarily who we ACTUALLY voted for.
True in general, at the presidential level anyway: you get who the electoral college elects, not who the voters elect.
Not true in the specific 2020 case, however: the electoral college selected the same person who the voter selected.
(Note that even the conspiracy theorists who say the election was stolen don't disagree with that: They are asserting (without evidence) that enough votes were "stolen" that the electoral college balance flipped... but it flipped so as to make the electoral college vote go for t
Re: (Score:2)
Damned if they do, damned if they don't.
China gets criticised for not releasing data early enough and for anonymising and aggregating patient data to protect privacy (because if the CCP does it then it's to cover something up).
On the other hand if your country happens to be seeing a pandemic you don't want any of your data being shared.
Re: (Score:2)
Re: (Score:1)
Most people I've seen wearing a mask -- they flip the mask up when they need to sneeze
Yep I have seen that also, I sneeze right into a mask while also pressed against a sleeve - but that is why you bring extra masks, you just change into a new one.
Re: (Score:2, Funny)
Re: (Score:2, Insightful)
The data is telling us pretty conclusively that mask mandates have not and do not help
The impression I get, having spent time in several different states and both urban and rural areas over the course of the pandemic, is that mask mandates don't really affect whether or not people wear masks. Not surprising they don't work if they don't cause any change in the % of people masking up. There is, unfortunately, much less data available on _actual_ prevalence of mask-wearing.
Re: (Score:2)
Re: (Score:1)
Even before the newest versions of the monstrosity appeared, it was well known that most masks are far more effective as egress filtering than ingress filtering for the obvious reason: Particles expelled from the 100% humidity of the respiratory tract shrink, and the smaller the particle the harder it is to filter (it gets easier below 300nm, but breathed out particles can't generally get that small because of their non-water content).
With something as malignant at this
Re: (Score:2)
What I have a lot of doubt around is that the masks most of the populace actually wear, have helped even slightly. Look around at any kind of comparison you like, and you see again and again that mask mandates have done exactly zero to improve results.
For COVID, the virus can enter through your eyes too, so you need a shield in addition to a mask if you want to be safe.
Re: (Score:2)
Why are you coughing in people's faces?
Re: (Score:2)
Why are you coughing in people's faces?
Only in the faces of people I don't like, of course.
More seriously, you should know that when you speak, a lot of spittle comes out of your mouth. Even if you are a sanitary person.
Re: (Score:2)
What about kissing? I heard that exchanges spittle too.
I strongly suggest not kissing everyone you speak to. That is something I refrain from.
Re: (Score:1)
Which is the reason you wear a mask. So that the spit that would fly out of your mouth is getting caught in the fibers of the mask. Most of it anyway.
It is my impression that people think they wear masks for their own protection. Instead, the masks are to hopefully prevent you from spreading the disease, but more likely spread it is very tiny doses that are not enough to infect others outright.
Slowing the disease down is the goal, and for that purpose, masks work well. Quality of masks varies a lot though.
Re: (Score:2)
Well, I am happy to wear a mask outside, just to stop the flu and cold.
I do think that mask wearing should still be obligatory for quite a while still, until that Delta variant has run its course on the unvaccinated.
What is the goal here? Since unvaccinated people are basically all going to catch it either way.
Re: (Score:2)
A mask lowers the 'viral load'. Getting a high viral load can result in more severe symptoms and increases the likelihood of hospitalisation and death. Conversely, if you get a small enough viral load, your body may be able to deal with the virus with little or no symptoms much like a vaccine.
Re: Just wait until someone carefully studies mask (Score:2)
The virus spreads through respiratory droplets, which you produce through regular breathing and speaking. You don't (typically) get sick from someone sneezing at you, but from spending ~15 minutes in the same enclosed space as them.
That was one of the things medical authorities got terribly wrong at the outset, but fortunately caught onto, at the same time clearing up a long-established medical misconception about the size of aerosols [wired.com].
Re: (Score:1)
Re: (Score:2)
What I have a lot of doubt around is that the masks most of the populace actually wear, have helped even slightly. Look around at any kind of comparison you like, and you see again and again that mask mandates have done exactly zero to improve results.
For COVID, the virus can enter through your eyes too, so you need a shield in addition to a mask if you want to be safe.
Depends on your threshold of "safe". Bear in mind that getting infected requires getting exposed to a decent number of virus particles, and that your eyes don't have large volumes of air passing across them most of the time.
So although it is possible to be infected through the eyes, most people's eyeglasses and sunglasses already get them close enough that it probably isn't worth the effort to do much more than that. Just a gut feeling.
Re:Just wait until someone carefully studies masks (Score:5, Insightful)
They've been carefully studying masks since day 1, that's why the recommendation to not wear non-medical grade masks got reversed, because evidence started showing that they reduced spread of COVID and the evidence has only grown since then.
From Florida which dropped mask mandates early to arrive at about the same results as mask-mad California,. to Sweeden which has has nearly zero deaths in 40 days now from Covid and never had a mask mandate at any point. The data is telling us pretty conclusively that mask mandates have not and do not help, I just wish someone would put this to a serious scientific study so actively harmful re-masking policies could end.
Fortunately people have [healthaffairs.org] done more [pnas.org] than a few cherry picked examples.
The reason why re-masking is actively harmful is that it greatly diminished faith in a vaccine that is otherwise shown to be working extroniarily well. The direction the U.S. Should have gone to improve vaccinate rates, is after the Provincetown study say "wow these vaccines are working so well, our recommendation is that at not time does any vaccinated person ever need to wear a mask".
Yeah... that sounds like a really unpersuasive line of argument.
People who aren't getting vaccinated are exactly the people who are going to stop wearing masks at the first opportunity. Which is actually a really good argument for continued mask mandates, because it's the only way to stop the dumbasses who won't get vaccinated from going around maskless and keeping the pandemic going.
Re: (Score:3)
They've been carefully studying masks since day 1, that's why the recommendation to not wear non-medical grade masks got reversed, because evidence started showing that they reduced spread of COVID and the evidence has only grown since then.
Not quite, but close. The recommendations around medical grade masks were based on the available supply of medical grade masks. They were always known to reduce the spread of infection. The ability for people to absorb and understand instructions is why this appeared as a reversal. Inter-agency and inter-governmental the instructions were to prioritise medical masks for medical professionals. Too hard to understand so the CDC guidance: don't buy masks. When enough became available that guidance changed. It
Re: (Score:2)
Do you believe the recommendations Anthony Fauci made to his personal friends via email were intended to preserve medical grade masks for health care professionals?
What a bunch of friends do personally is completely irrelevant when talking about global government policy, and quite critically: is not actual policy.
Re: (Score:1)
From the paper you linked, near the end, it reads to me like a casual but deliberate brush off the importance of missing study/info:
"The impact of using masks to control transmission in the workplace has not been well studied. One issue that impacts both school and work usage is that, over a full day’s use, masks may become wet, or dirty. A study of mask use in health care settings found that “respiratory pathogens on the outer surface of the used medical masks may result in self-contamination,
Re: (Score:3, Informative)
The evidence has shown only the opposite, as I already explained from comparing regions with different mask mandates.
No, it doesn't. What you showed was the difference between those who wore masks and those who did not in comparing regions. It has been demonstrably shown that in those places where masks were mandated and use was highest, covid rates significantly declined [healthaffairs.org].
There was a significant decline in daily COVID-19 growth rate after the mandating of face covers in public, with the effect increasing over time after the orders were signed.
This data was also backed up two months later [nih.gov].
Conclusion: The study suggests that community mask use by well people could be beneficial, particularly for COVID-19, where transmission may be pre-symptomatic. The studies of masks as source control also suggest a benefit, and may be important during the COVID-19 pandemic in universal community face mask use as well as in health care settings. Trials in healthcare workers support the use of respirators continuously during a shift. This may prevent health worker infections and deaths from COVID-19, as aerosolisation in the hospital setting has been documented.
Which is exactly what this visualization shows [livescience.com]. Wearing a mask significantly reduces par
Re: (Score:1)
because evidence started showing that they reduced spread of COVID and the evidence has only grown since then.
The evidence has shown only the opposite, as I already explained from comparing regions with different mask mandates.
Your link is frankly a pretty stupid thing to bring up since we now have actual evidential results from incredibly large scale trials, rather than extremely limited studies. It's to be expected now though since the worship of masks has reached a cult-like status now and people will grasp at pretty much anything to keep them in use.
As I said, we have clear proof
Clear proof?!?!? Are you kidding me?? All you did was list 3 jurisdictions with different mask policies. You don't even know the relationship of mask mandate vs actual mask wearing, not to mention really essential factors like mask prevalence vs local spread, and how the causality of that relationship plays out, or a dozen other factors you need to consider when doing an actual study or two, like I posted, rather than the jazz hands that seem to count as persuasive evidence for you.
Re: (Score:2)
The evidence has shown only the opposite
No it hasn't. The WHO extensively commissioned studies on this and the evidence was clear. Not only do medical masks help, but they updated their guidance to include home made masks as well.
But feel free to point to me the giant aggregate study you did that contradicts the recommendation from actual people in the field.
But you won't. Since the very start of the pandemic you have been one of the biggest peddlers of bullshit on Slashdot about masks (at least one of the biggest not posting AC, but I suspect yo
Re: (Score:1)
That certainly isn't true. Normal people do not like wearing masks and thus will get rid of them as soon as possible. It is the virtue signalers who won't.
Normal people will do what public health authorities suggest as the best way to protect their fellow citizens. Antisocial assholes who don't care about others won't.
FTFY
Re: (Score:3)
Nope it has nothing to do with virtue signaling but logic. Masks reduce the number of virus particles you inhale or, if you are a carrier, exhale to be widely airborne.
1. A reduced number of virus inhaled particles means less chance of one virus getting through to a cell where it can replicate.
2. If it does get to a cell that can replicate, given that the doubling time is 9 hours the fewer viruses your infection starts with, the longer it will take to ramp up to massive enough numbers to cause major damage.
Re: (Score:2)
Based on the number of typos in your post, you might want to consult with your doctor to see if the vaccine is just as extroniarily as you claim.
Re: (Score:3)
From Florida which dropped mask mandates early to arrive at about the same results as mask-mad California,. to Sweeden which has has nearly zero deaths in 40 days now from Covid and never had a mask mandate at any point. The data is telling us pretty conclusively that mask mandates have not and do not help, I just wish someone would put this to a serious scientific study so actively harmful re-masking policies could end.
There have been some studies here, and the careful studies strongly suggest that masks help. One of the better studies here is this one in Germany. https://www.pnas.org/content/117/51/32293 [pnas.org]. Since different parts of Germany had mask mandates at different times, they were able to compare COVID spread rates in very similar regions at the same point in the pandemic. Since Germany is also pretty homogenous compared to different entire countries, this also eliminates some of the complications from that. (Germany
Re:Just wait until someone carefully studies masks (Score:5, Funny)
What will be great to see in probably about five years time, is a series examination of masks and if they help.
It has been done. They help. A lot.
I have no doubt if you are sick masks help prevent others from being as exposed. I also have no doubt hospital grade masks, well fitted, can help prevent exposure.
And that's really the point. COVID is contagious about a day before you're sick. Even crappy masks, poorly fitted, can dramatically reduce exposure if they're right at the source where all of the virus-encumbered air is coming out, while the droplets are still big.
What I have a lot of doubt around is that the masks most of the populace actually wear, have helped even slightly. Look around at any kind of comparison you like, and you see again and again that mask mandates have done exactly zero to improve results.
No, that's exactly the opposite of what literally every comparison has shown.
From Florida which dropped mask mandates early to arrive at about the same results as mask-mad California,
When you're talking about millions of people, Florida having 18% higher case count per capita is statistically significant, particularly when you consider how much of southern California is every bit as anti-mask as Florida. If you ignore the right-leaning parts of the state and just look at the Bay Area, the numbers look even more radically different:
That's about half as many cases in per capita in the Bay Area as in Florida. The difference? Everybody wears masks, almost everybody works from home if they can, there's a much higher rate of food and grocery delivery, so people aren't out shopping as much, schools are mostly remote, etc. Those things all add up.
to Sweeden which has has nearly zero deaths in 40 days now from Covid and never had a mask mandate at any point.
And has very low genetic diversity, with fairly similar genes when compared with neighboring countries that mandated masks, and a much higher death rate than any of those neighboring countries.
You can't just blindly compare numbers. You have to actually understand what the numbers mean if you want to get any useful information out of it.
The data is telling us pretty conclusively that mask mandates have not and do not help, I just wish someone would put this to a serious scientific study so actively harmful re-masking policies could end.
False. Literally every single scientific study that properly controlled for population differences has shown exactly the opposite of what you're saying. These studies have been done in multiple ways, with the most striking differences shown by comparing counties with mask mandates against adjacent counties without mask mandates. Those studies showed that passing a mask mandate resulted in case counts dropping while the case counts in those neighboring counties skyrocketed.
The reason why re-masking is actively harmful is that it greatly diminished faith in a vaccine that is otherwise shown to be working extroniarily well.
Extraordinarily well? No. About as well as expected? Almost. As well as is likely needed to stop the pandemic? Maybe, maybe not. You're right, though, that there's probably some percentage of people who will misinterpret the reintroduction of mask mandates as an excuse to not get vaccinated, though it is doubtful that many of them would have gotten vaccinated anyway. Realistically, if you want to be vaccinated, you already are.
The direction the U.S. Should have gone to improve vaccinate rates, is after the Provincetown study say "wow th
Re: (Score:2)
Is the Funny mod interested as censorship? Or a backhanded way to give you more visibility?
Most of the points seem valid, but arguing with a fool? "That trick never works."
Re: (Score:2)
I have a different perspective on this. If nobody responds to posts l like that, there's a real chance that somebody might believe that people are censoring the truth. I'd rather spend a whole lot of time tearing down bad arguments than risk some innocent person stumbling across them, getting the wrong idea, and being permanently disabled or killed as a result.
If this were about something minor, I might let it go, but this subject is too important to let a post that's so misleading be the last word, even
Re: (Score:2)
In my previous reply, s/interested/intended/ (because my fingers sometimes pick their own favorite words?).
On the substance, I agree with you about the importance of the topic, but disagree with you about the tactics on two counts. Also, I have a secondary idea for a solution approach involving masks.
One tactical problem is that narrow targeting at a strong target is usually ineffective. The best attacks go for weak points, but in this case the thick skull of a troll, possibly a professional troll, is a kin
Cures and wonder drugs (Score:2)
Sure sounds a lot like snake oil or a number of other unregulated 'cures' and pharmaceutical-like products. Lots of claims and little proof.
Bad data not useless AI (Score:1)
Re: (Score:2)
It is mostly bad and insufficient data and not necessarily AI (Machine Learning) being uselsess.
If you don't have the data then the AI is useless.
It reminds me of all the contract tracing apps. If you decide you want to solve COVID with an app then contract tracing is it, but the people trying to solve COVID were really wasting their time (and precious public attention) promoting contract tracing apps.
Same with AI. If you want to use AI to solve COVID then sure, classification on X-rays or physician notes is the way to go. But I haven't heard of any big wins involving AI and other medical conditions,
Re: (Score:2)
AI iss good at photos and audio (Score:3)
If you give an AI a photo such as an x-ray, it can look for tiny patterns. Similar for an audio recording. These have LOTS of data to go deep into. Things too small for a Doctor to notice casually.
But for Covid and similar things it is analyzing things like Doctor's notes on what patients answer.
In effect, you are only analyzing the things the Doctor's notice, which are minimal data.
So you only recognize Covid if the Doctor did.
"Built to Catch Covid" (Score:1)
My AI bot caught Covid. Was in the eHospital for several weeks, and its wanker no longer works.
Given how Covid was running, it's fair to say (Score:2)
Artificial Intelligence is no match for human stupidity.
Re: (Score:1)
We could have wrapped this thing up in 6 months or less, before a vaccine even. If only we didn't spent the first 14 months in denial. I mean honestly, if China shit the bed then telling everyone isn't going to get the sheets changed. When I was a young Republican voter and talk show junkie, I used to think conservatives believed in rolling up their sleeves and getting the hard stuff done, now I think they're a joke.
We gave it our best shot... (Score:2)
blockchain: https://www.ncbi.nlm.nih.gov/p... [nih.gov]
Re: (Score:2)
Obviously, it needs more cowbell ... er, I mean blockchain.
Actually a dual failure (Score:3)
First, it's a failure on the level most here are noting - a medical failure. Apps not helping in many ways related to tracking, diagnosing, treating etc an illness.
Second, however, is a failure I actually consider more large and important. Here in the 2020s altogether too many people are addicted to cell phones and too many tech types are addicted to solving things with a little scripting, or some Python or Java, etc. Not all problems are solvable with a stupid app running on some shiny portable object!. Sometimes, people must take real world actions in the real physical world to actually solve a problem, but too many people these days want to sit in an air conditioned office and hammer a keyboard and then claim to have done something to solve a problem - and it never crosses their minds to put down the tech and put on some shoes and some gloves and step out to do something concrete. It's like some new form of virtue signalling. Imagine an alternate history where the child asks grandpa "what did you do in the war?" and the response is "I wrote an app!".... very sad.
It's often been said that when you give a kid a hammer, the kid thinks the whole world looks is made of nails - we all need to step back, take a deep breath, and accept that the whole world is not fixable with an app.
Well, Artificial Ignorance cannot really do much (Score:3)
All it can, given a _good_ training set is badly, but cheaply copy what expert can do.
There seems to be this misconception that AI can do more or do things better than human experts. That is fundamentally wrong. First, there is a lot of things that present AI (and that includes all known scientifically sound theoretical models as well) will never be able to do. In that real, anything requiring insight or understanding falls. AI is as dumb as bread. It has absolutely no understanding about anything. All it can do is fake things. Second, faking it can work for a lot of applications, but it will always be worse than when a real expert does it. Now, there are quite a few application areas where the people doing it are routinely not experts, for example driving. There are quite a few application areas where the level of skill needed is actually very low. These are areas where AI can make a difference. Most of what it will do there is eliminate low-skill jobs though, i.e. jobs of people that can really not up-train.
In short, expecting AI to make a difference in anything new or not very well understood is foolish.
The algorithms are only as good as the user (Score:1)
Data quality has been poor (Score:2)
Meanwhile HIPAA makes it difficult to associate COVID cases with other conditions and treatments. I think this has been the biggest barrier to consistent data sets - p
I realized this about last June... (Score:1)
Then it was July.
Then it was August. And still no release.
And then I sat and sat as the case count turned vertical that godawful winter, wondering to myself, "Obviously it's not 95% in real life but even 65% would be a godsend for being able to punch ba
What assumptions were they making? (Score:2)
You can't start with no information. Hopefully, you start with pure, unbiased observations and not just a few but all of them without cherry-picking. I'm guessing that that didn't happen and the programmers made some assumptions. What were they?
Not a surprise (Score:2)
AI/ML, while being seriously interesting, are the new buzzwords. I'm not surprised to see these solutions being pitched where they either make no sense or are something simpler branded as AI. Blockchain, AI, ML. You can't sell a toasted sandwich in California without claiming it 'utilises' one of those technologies.
Having worked in health care ... (Score:2)
... the data is crap because the medical system in general has no clue how to collect it nor any inclination (in general) to do so. Here's the situation:
Data is an after-thought - the mentality is almost exclusively on the here-and-now. While good for an emergency situation, it means the data collected is usually relevant to _post_ treatment. Data is rarely collected until after something has already gone wrong and been addressed. That's why there's no data for early detection ... nobody collects data
Da
ML objectives need to change (Score:1)
Re: (Score:1)
Re: (Score:1)
Re: (Score:1)
Re: (Score:1)
Re: (Score:1)
I was watching the Olympics. The same world-class athletes who just competed side by side without masks, probably breathing heavy from strenuous athletic activity, were required to wear the face diapers on the podium of 3 people. Immediately after the anthem was played, they took their masks off to take pictures. No media outlet or anyone said a word. At this point, masks are just virtue signaling.
It's about establishing a norm of mask-wearing except when otherwise necessary.
Re: (Score:1)