Under AI’s Watchful Eye, China Wants to Raise Smarter Students (2019) [video] | Hacker News Reader

Under AI’s Watchful Eye, China Wants to Raise Smarter Students (2019) [video]

wsj.com

53 pointsseventyhorses46 comments

Everyone's focusing on the pressure this adds but that seems largely due to cultural factors. Of course students feel more pressured if the data is primarily used to criticize them. But what about evaluating the teachers ability to keep students focused on a task? Or evaluating if a new medication is making a difference for a student with ADHD? As a side note, it seems like lazy journalism to not seek out more commentary from people involved in the project. Their expert was a random Neuroscientist from San Fransisco for gods sake, they could at least have explained why they found his particular criticism worthwhile but they never even establish if he's familiar with the specific work being done in China. If the research is being done in China and the tech is bespoke to the research program in China it's weird to prioritize input from American researchers without some degree of explanation is all I'm saying.

> If the research is being done in China and the tech is bespoke to the research program in China

The headbands are produced and donated to a Chinese school by an American company! The people protested and the authorities told the school to stop using them.

The headbands were handled via a partnership with Zhejiang Brainco Technology Co Ltd. so donation is far from the right word to use. One of their partner's cofounders is also a Harvard PhD holder so it's hardly a case of the Chinese company just being there to take up space as is often claimed around here. This is a donation in the same way that lots of large American companies are Irish. And the objections from parents as well as the government's cited reason for the shutdown were on data privacy grounds according to multiple articles which is a valid criticism but unrelated to the actual usefulness of the devices. So I'm not sure what you're actually driving at here.

> The headbands were handled via a partnership with Zhejiang Brainco Technology Co Ltd

Which I guess is a subsidiary of the US-based company BrainCo located in Massachusetts. The article describes the provisioning of the headbands as a donation, which I guess makes sense if you're a startup and you want to test a pilot program in a market that might be receptive to your product.

The fact that Chinese parents protest on social media and with the authorities (on ethical grounds, it appears from the Quartz article) and the authorities banned the device on privacy concerns, sounds a lot like what would happen in the West.

As an aside, every time I read an article about the Chinese society I discover a much more nuanced picture than "it's a dictatorship with no free press and human rights". It's not a democracy for sure, but the idea it's a totalitarian state where all dissent is banned seems very far from the truth.

> Which I guess is a subsidiary of the US-based company BrainCo located in Massachusetts.

Why am I not surprised. A U.S. company designs a device that likely does not work but can be used to control people who think it works.

No orders come in, so let's donate it to a Chinese school for a pilot project in the hope that the CCP bites.

> every time I read an article about the Chinese society I discover a much more nuanced picture than "it's a dictatorship with no free press and human rights".

This is a generalizable realization. If you’re aligned with any political faction, it’s best to assume that what you hear about other factions is more negative than the reality.

“ a much more nuanced picture than "it's a dictatorship with no free press and human rights"“

You should assume this is the case about everything that you have only read about without having long time personal experience. What we hear in the media about other countries or times in history is often basically a caricature of a reality that is way more complex and subtle than we usually are told. This is certainly true for China but also true for things like the Nazis or living under communism.

Reality is way more complex than we have time to learn.

as an autodidac that did extremely poorly in school, but flourished when people FINALLY left me alone and let me concentrate on what I am interested in: I could not imagine what kind of hell this would be. For a lot of the population School sucks and is a waste of time.

I have the exact same feeling. I'm not sure how well these will work, but if the person isn't comfortable, it only leads to mental health issues later on on life.

School is tough enough to begin with. I can't imagine how much more difficult and stressful it will be for these kids.

Obviously in this case the mental health of these students is tragically coming in second place.

Beyond that it's also naive. I can't think of a faster way to destroy a work force's future variance in creativity than by having them all optimized for an opaque algorithm.

Many of the world's most accomplished found their genius through events and choices that cannot be optimized for. Heart break, substance abuse, loss - to name a few.

What did you succeed in?

Totally agree; I left high school early (no diploma) so that I could go to college. In high school I hated it, didn't do my homework, got poor grades, hated my math classes, and was generally inattentive; but I loved college, breezed through most classes, and discovered that I loved mathematics, so I majored in it, and now I've started a successful company.

My experience is different than most peoples', but the key thing is that people have a _choice_ and can find out what works for them. The more we try to fit people into algorithms, and top-down control, the less they can go find out what they are happy doing and how they want to do it. Control, coercion, surveillance and manipulation tend to choke out the beauty of life.

I’m similar to you - autodidact, poor in school, successful later - and I think it depends how the A.I. is implemented.

My observation of school was that the ratio of 1 teacher to 25 students (who all have different learning speeds) means that the teacher is optimally teaching only a fraction of the kids, the rest at each side are overwhelmed or bored.

If the AI is able to help the slower students and stimulate the brighter ones, then I think that would have taken away a lot of the boredom for me and made it a much better experience

>the ratio of 1 teacher to 25

As an aside, I've often wondered if it's not class size that's the issue, but conflicting ability and personality types. Things like the GATE program grouped high ability kids, but personality matters and it would be interesting to match students together and with a compatible teacher. Of course, this might fail spectacularly as one of the benefits of diversity is getting other perspectives, and learning empathy. But I think it would be an interesting experiment, especially in a class like math.

Class size impacts at all levels up to college. I think teachers will be the first to acknowledge that they enjoy classes with a a dozen or less students.

I had good teachers who just let me sit at the back of the class and work at my own pace, they would mark any work that I handed in.

Same feeling here. My gpa in college was absolutely abysmal, but I've since sold a startup, and working on deep tech right now. School really measures nothing but how good you are at adhering to protocol.

Training kids like this reminds me of training AI itself:

Make the reward getting a better grade and the kid will only optimize for that. Eventually the kid (the agent) will be out in the wild and they will have a weak understanding of what the reward is and how to optimize for it (it will be harder for a fine-tuned kid to do domain adaptation towards different and more abstract fields and problems).

There’s another side to this issue too: Professors themselves are often optimizing their content for testability, rather than for providing actual, functional lifelong knowledge.

> Professors themselves are often optimizing their content for testability, rather than for providing actual, functional lifelong knowledge.

Hah! I think you have this backwards. Most professors are quite concerned about imparting knowledge when the vast majority of the students simply want a piece of paper with the University's logo on it.

I knew of very few professors in electrical engineering or computer science who tried to optimize for "testability".

The choice of using the word professor, seems misleading to me. Far more often teachers will adjust their content to ensure optimal performance in tests. Especially in a education system like this.

I can only imagine what side effects this must have for motivation. When you are cautiously trying to appear focused, can you even develop a passion for a subject?

Oh, you mean high school.

In the US, at least, the answer is actually "Yes. You can even develop a passion for a subject and instill it into your students."

My father taught English almost 4 decades in a very rural school. No Child Left Behind had him quite concerned intially.

Until it came down ...

He laughed at it. "Flag poor students early"--okay, his grades were in Excel since he was terrible at math--sort, print, tell administrators to go away. "Need writing samples"--"I have two entire filing cabinets of writing assignments for this years classes. Which drawers do you want?"

Now, I have to avoid generalizing. My father did not teach in an inner city school (the area had brutal poverty--but probably less crime) where they were continuously under threat of being forced closed by the government. He also did not teach in China or South Korea where getting the piece of paper seems even worse than the US.

However, the US isn't as brutal on the "teach to the test" fronts as many other places.

Here's a more complete story:

https://qz.com/1742279/a-mind-reading-headband-is-facing-bac...

Seems that the headbands were donated to a school by a US startup that put them on the market as a consumer product, probably targeting China's 'tiger' mothers; that people protested, and the authorities banned them.

The video says "this school is using" and while it sounds like an generic example, it's actually ONE school, using for a short period of time a product donated by a US startup.

I remember reading this in 2019 even before I ever noticed(2019) is a part of the title. The tech antagonized me by a lot, to my belief, the attempt went way way beyond necessary evil and also scented of scam. This is disgusting, it's narrated by a curious young male voice, explaining how the device is working on himself, his classmates, and how the teachers and others will know then their attention is drifting away from the class. This shouldn't happen, I can't even imagine how they got their IRB approval in the first place. Collecting and using EEG data from a consented subject for a scientific investigation is one thing, but using a poorly designed toy to regulate human behavior is another. The former one has an explicit scope that helps the human race to understand itself, but the later one is an onset of nefarious control freaks en masse that targeting not only your body, but also your mind.

Btw as a previous research assistant who spent 5+ years in brain imaging, the tech behind this also looks weird to me, you can't get meaningful EEG data in a room with so much interference. I don't see any follow-ups on this either. It usually takes 2 painful hours, give or take, to properly prepare ONE subject for an EEG experiment, and 15 minutes at least after that for them to wash off the glues on their head.

It reminded me of Brad Pitt's character in "Ad Astra", when he talks about "compartmentalization". As a human, you become part of a machine that is society / the military. You become part of "an effective team".

This is probably just research from some institution. Why the WSJ generalize and say that "Under AI’s Watchful Eye, ---> China <--- Wants to Raise Smarter Students".

They should rephrase it into something like "Research tries to find if..."

I trust western media the same as Chinese media...

(Edit: format)

> media the same as Chinese media

Not quite. Western media lies because they want to earn money and survive.

Chinese media lies because they are government controlled and the CCP propaganda arm has a well planned communication strategy to steer public opinion in the desired direction. Both in China and worldwide.

The design is ironically could be seen as an anology to one of the most famous novels in Chinese history of all time - 'The Journey to the West'[0].

In the novel, the monk Sanzang has a golden ring on one of his disciples - Wukong's head, and it was used for causing extreme headaches with a spell, in order to control Wukong to follow his will because he is smart but violent, and IMO sounds like an ADHD guy.

[0]: https://en.wikipedia.org/wiki/Journey_to_the_West

The primarly role if these headbands since more like conditioning to me than to actually read brainwaves.

You condition them as kids to respond to these headbands, then sooner or later companies will require them for workers.

Yeah, the headbands appear similar to Scientology's "E-meter", which is a fake tech device used to control the victims.

It seems "1984" and "Brave New World" came with a bit of delay. First rule of abstraction, whenever you abstract something you loose intrinsic value. Why is "concentration" favouritised? What about other brain states. I am afraid this experiment is wrong as much as it can get, what we believe we become.

So I can easily imagine, kids that deals with emotional issues of tyrant parents, and then get double punished because they were dealing with those emotions.

I thought concentration was weird too. I've had a few EKGs and they don't read anything above the noise when I'm doing simple math (which is what they ask most people to do). So I just started coming up with really difficult problems for myself to do and bingo, they can read activity again. This leads me to believe that if someone is performing well but concentration is in the noise or low, that you would want to instead push them to new topics, not punish them. Giving this data to parents seems like a bad idea because of this. You already know material well, so while the teacher is discussing it you drift off or your brain doesn't have to work hard to take information in. Parents see low concentration so get mad because you aren't using your time in school wisely (which is true, but not your fault). You then score low on tests because of the stress. Cycle repeats.

In addition to grades, these headband readings add additional pressure to kids. Forcing kids to improve their grades at the cost of their mental health is unacceptable.

> Forcing kids to improve their grades at the cost of their mental health

This sounds like the status quo of many educational systems worldwide. It's very depressing to me. Obviously some countries are worse than others, but even the US isn't great compared to say, Nordic countries.

This.

It's easier to criticise other countries' education systems, but for some reason we try to avoid criticising our own.

To be honest, few today are educated. The most favorable characterization is that we're putting the youth though a mill that produces semi-technically savvy savages. That fact that so few people with diplomas (of which I am one) fail to see that is because they haven't the slightest clue what being educated means. Also, fallacy of sunk costs (that piece of paper was rather expensive and time consuming).

I wonder if there will be some sort of feedback where they discover the conditions under which kids ACTUALLY learn and thrive.

Like they figure out the successful ones slept in every day, had sex early, went to parties and drank from time to time.

On the other hand, all school teachers pay attention to check that their students are focusing. They use neural networks for that (their own neural networks) that are very well trained in detecting lack of focus from body movements, facial expressions, chatter, etc. A device that actually allows more easily to detect focus is scary if it's used improperly, that is, if focus is required all the time. But if used in a balanced way allowing for prolonged periods of relax and "mind wandering" it might actually be helpful.

This stuff is SciFi and doesn't exist.

FFS if it exists sell it to adults and make $$$

This conspiracy that China has all this secret technology is a bit embarassing to believe.

Sounds to me like a fashist's wet dream.

Because micromanaging people is how you get more Feynmans and Shannons.

Bengio, Hinton, and Lecun were pariahs for decades, until all of the sudden they were rock stars.

I feel so bad for the 1.3bn people living under this. If we're not careful its going to keep creeping into the West and creativity will suffer.

Edit: removed profanity

> I feel so bad for the 1.3bn people living under this bullshit.

So the WSJ succeeded in its purpose of spreading anti-Chinese propaganda. Wow, so easy. Check a more complete story here:

https://qz.com/1742279/a-mind-reading-headband-is-facing-bac...

EDIT: less snarky. Please pay attention, I have the feeling that Western media are building an anti-China sentiment because the West (the US in particular) feels threatened by China's economic and technological growth. Criticism of China is fine and due, but let's avoid being tricked by the media (of both sides) in re-aligning with our own tribes. This is the way conflicts are prepared and started. We need understanding, diplomacy and cooperation, not anger and sabre-rattling.

Ever heard of Agile?

>Bengio, Hinton, and Lecun were pariahs for decades

But not as much as Schmidhuber.

Reminds me of a certain novel I read about few years back where certain people would become living calculators through some kind of chemical intervention. Losing all personality in the process they'd become incredibly productive savants capable of sifting through an ever increasing volume of data.

Guessing you're also excited for December 18? ;)

Reply on news.ycombinator.com