Ask HN: AI “white mirror” – what are some inspiring future visions you've seen?

52 points by tikkun ↗ HN
We have black mirror. But no white mirror, yet (Microsoft/OpenAI, this could be a cool thing to do ).

What are some of the best inspiring visions of the future you've seen?

Both distant ones, as well as ones that could be possible within ~1-2 years

80 comments

[ 3.0 ms ] story [ 164 ms ] thread
AI babysitter for infants to supplement – not replace – parents. Captures audio / biometrics and responds to the baby with targeted music, positive affirmations, cooing, etc, to promote positive emotional state. Also notifies parents of states that require their intervention, like hunger, soiled diaper, abnormal vitals, inconsolability. Parents can take a break, watch a movie, entertain friends, etc, with the comfort of knowing their baby is being intelligently monitored. Existing baby monitors are way more noise than signal.
That's 100% black mirror, not white mirror. Babies being raised by machines is dystopian as hell.
Baby isn't being raised by a machine. How is this different from a crib mobile (for entertainment) + video monitor?
If it's the same as a crib mobile, then what benefit does it provide?
More effective
The effectiveness comes from exploiting the baby's primal recognition of the mother, which is why it's dystopian. Additionally, by "effective" you mean "the baby spends less time crying," which means that the parents are spending less time with their baby. So parental time is being replaced by an AI. I think it's fair to call that "being raised by AI."
I hope the "Robot nanny optimizes for minimal crytime" argument wasn't written by a parent. What a cold angle.
(comment deleted)
There's some point where the robot is realistic enough that it's functionally no different than a nanny. I agree it would be dystopian as hell if the kid was just raised by the robot, but giving the parents a bit more sleep and an opportunity to have a date night once a week would help them be way better parents.

As an extreme case, some parents get so stressed and chronically sleep deprived that they get frustrated at their babies for continually crying and shake them, leading to permanent brain damage and even death.[0] To be clear, these parents don't mean to hurt their child but they're not operating coherently anymore. A robot that can give parents a breather here and there would significantly improve quality of care for soooo many babies.

[0] https://www.mayoclinic.org/diseases-conditions/shaken-baby-s...

Already a thing. The machines are called smartphone/YT/TT/IG, etc
Batteries die and the parents are left without a clue how their child actually operates. They've been playing real life tamagotchi while their newborn evolves into an alien.

Funny, babies are also way more noise than signal >.<

"You won't always have a calculator in your pocket"
I think most new parents have semi-wished for, like, a life-size cutout of themselves attached to a Roomba that could roll into their nursery at 3am when the baby randomly wakes and make vaguely reassuring noises to soothe them back to sleep. I know I have. But I'm not sure it's wise to enable or encourage this level of parental disengagement.

Anyway, what I really wanted to share is "The Robot and the Baby", by John McCarthy (yeah, the same one): http://jmc.stanford.edu/articles/robotandbaby/robotandbaby.p...

Selling it as "free up the parents to do other things" is not going to fly. But enhancing the baby's experience or learning is fine.

An AI-powered rocker that plays soothing music till the baby sleeps would be great. People already use rockers, it's very accepted.

I would never have trusted a computer to be in charge of my infants. Even trusting another human adult is pretty tough already.

The Control Problem is scary enough already. What's the Paperclip Galaxy equivalent for a robot nanny?
While that's decent, I would wish for something more like, "A boy and his dog" type AI. Having an "invisible friend" that you can speak with and see, who knows everything, is a perfect confidant, who grows with you through life until you are ready to set them down, possibly as a rite of passage into adulthood.

And even then, you could keep in in your computer or phone, be able to talk to it and discuss problems with an intelligence that knows your life almost as closely as you do, but not carrying it with you all of the time.

IDK, I think that would be pretty cool even if the idea of robot nanny scares the pants off of some people.

I anticipate AI will slowly start to be realized to reveal the nature of not only human language, but also human cognition, which may lead to us starting to become more self-aware as individuals and a species......like, if we happened to somehow have accurate historical statistics on the matter, we'd see major jumps on the chart representing significant events in history (invention of the printing press, the first enlightenment and rise of science / retreat of religion, etc), and this event will (possibly) turn out to be another instance of the relatively rare phenomenon.
I think AI-assisted healthcare is inevitable and will probably offer the biggest improvement in the general population's health for decades, it will focus on disease prevention and flagging up those early warning signs that most people don't usually notice
10 years ago I wrote my academic thesis claiming electronic medical records would fix these things too. They won't.

An early warning system is useless if there isn't any funding for public health programs that could help people address said warnings.

> funding for public health programs

Why do you believe the solution needs to come from tax collected revenue? If you had a way to make people live healthier, don't you think people would pay you for it?

Unfortunately people often don't go to the doctor because it's too expensive. Why would that change if you added more things to the doctor's toolbox?
What creates better public health outcomes, making condoms available widely available at no cost, or letting market forces take care of it?
Sure, the wealthy would have means to pay for cures, life-extension, organ-regeneration while the poor would not. Imagine the dystopia that creates.

Let's say there was a 100% effective treatment for cancer/MLS but cost $20,000 per month. We'd all be "willing to pay" for that for a loved one but "willing" isn't "able." That's the problem with US-style for-profit healthcare. Economics breaks down because there's no ceiling to what the "consumer" is willing to pay, only a profit-based decision by companies of how many people it wants to divide into people deserving of health and those not.

Seriously, as a society, if we don't grapple with economic ethics of "miracle cures", corporate greed plus technological advances will lead to endless class wars over who's grandma/child gets to live or who's withers and dies unnecessarily.

It'll get cancelled for fat shaming 0.2ms after being turned on
This would suggest some alternative "shortcuts" to good health for a broader population. Instead of the time-consuming approach which includes physical activity, fasting, good nutrition, etc and corresponding things for good mental health.
I think generative AI is going to created a modding and fanwork renaissance.

Mega game mods are seemingly bottlenecked by art, textures, voice acting and such, and fanworks outside of gaming have similar constraints. And the legal issues surrounding these models arent an issue for them.

AI service or feature that find common ground and rephrase concerns to help people find agreement.

In this contentious world we live in, I feel like all we really need is a little more love for our common earthbound inhabitants.

This could work for neighborly disagreements or even bigger stuff.

(More robotics, but if you'll accept it)

Workers of the dirtiest, least desirable jobs ( sewage, septic cleanup?) get paired with a robot that handles the dirty work.

No job loss, critical infrastructure remains, just less toil for the human.

(comment deleted)
It's not "black" because it's bad. It's black because it's a turned-off screen, which has a black colors. Black Mirror stories are bad because the Mirror reflects bad/broken people, not because the mirror is is bad. A White Mirror would be a screen that is turned on, but also reflects images.
Yeah op yikes to jumping to 'black = bad, white = good' metaphor space. What is wrong with you?
You're getting downvoted but I literally just got a memo at work last week reminding people to stop using the terms blacklist and whitelist and switch to block-list and allow-list.
(comment deleted)
Such a waste of energy and time to get upset at people using these words. Its not like using the word block-list and allow-list was the problem in the first place and using different words wont change anything.
There's a big difference between picking fights about traditional forms in language (wasteful) and pointing out that this person jumped to the wrong conclusion that black mirror = bad mirror, when it's just a poetic way of saying tv. Downvote, call me keyboard sjw be upsed I don't care. Big yikes from me OP, rethink things.
Sure but it’s clear what OP is asking. I always interpreted “black” as in mirror to be a double entendre anyway.
The one where we stop wishing AI or whatever other toy du jour will solve all our problems. The road to hell is paved with good intentions. I don't buy your utopia, no matter the label.

I guess my utopia is where people stop making utopias and start behaving pragmatically.

And off he went to search for berries!
Maybe this sounds a bit sad, but I'm personally super excited for when something like copilot turns into a full-on coversational code friend / assistant.

I see it being super productive to always be pair programming with an AI, and not have to type every command out... or better yet, that it provides suggestions without me needing to prompt it to do so, thereby giving this feeling that I'm just hanging out with someone and doing a more traditional pairing session

Perhaps a bit too distant, but the Culture series by Iain M. Banks takes place in a utopia built and managed by hyper intelligent benevolent artificial intelligences.
AI Financial management assistance, Healthcare price comparisons, Insurance, and cost saving systems - AI will be much better at navigating the complicated world of finance than the average person, will probably be a great improvement for most people
It'd have to be something like AI both solves climate problems while making it profitable to do so. I'm not talking about greenwashing, but using machine learning to solve hard science problems and tying them to economic benefit. Dear AI overlord, I'm trying to start a new business and this is my idea… Dear valued chat user, your business can be 125% more profitable if you can address this climate problem as these 3 VCs are ready to unload dump trucks of cash into the space.
In a distant sense, I find the Culture series of novels to be an inspiring vision of the future. People develop AI and technology to master nature on a planetary scale, and largely live peaceful, happy lives.

In a nearby sense, I am inspired every time I do random household chores, by thinking that maybe AI and robots will be able to do this soon. Folding clothes, picking up toys, putting things away where they go, cooking dinner, putting the dirty dishes into the dishwasher, putting the clean dishes away on the shelves, taking out the trash, driving to the corner store to pick up a fresh baguette.

I second this, AI in the Culture series is the most positive example I've read. It's notable that AI co-exist with humans everywhere in Banks' novels—unlike say Star Trek, where Data is a rare outlier.
Agree except for the cooking. Like gardening, its a very rewarding privilege if you don't have to so the other shit
Agree except for the cooking. Like gardening, its a very rewarding privilege if you don't have to do the other shit
To me this is the perfect answer.

Star Trek is good too. Particularly once you get into the ship computer and holodeck concepts. But Banks writes so specifically about about what it would be like for a civilization to thrive working together with strong ai.

The books together form a loose philosophy that is like Seinfeld for AI conversations.

very distant: The Culture series by Ian Banks Somewhat distant: I Robot by Asimov (specifically the ending)

I will not comment on the realism of either of these novels, but will say that both present optimistic visions of futures where super intelligent AGIs are commonplace and both have interesting ideas to consider regardless of exactly the extent to which they are grounded in reality.

Technology just magnifies what's already happening in society (occasionally it disrupts it). What are some good things in our modern society?

I'm not a biologist but probably mRNA stuff or medicine that works with your genetic code to give you true immunity, cure chronic diseases, etc.

Automation creating a post-scarcity environment where people focus more on expanding their freedoms and quality of life instead of working more and more (unlikely if the last half-century is anything to go by).

Social media fostering a sense of "global citizenship" making wars based on nationality less likely (or at least harder to drum up support for). Imagine trying to justify the Iraq war to US citizens in 2023. There would be Iraqi Youtubers making videos humanizing their struggle. You might have Iraqi online friends in your Discord servers etc who make you question if the US Military is actually a "force for good". We can see this happening in Ukraine as we speak, granted nobody was really a fan of Russia to begin with, but I think most of us can point to a Ukrainian influencer or someone in our social circle that is directly affected by the invasion.

I really love that last point. Being able to be closer to our fellow citizens of Earth really helps bring things down a notch. Being able to see the movement in HK and what is happening in Ukraine first hand has really affected how we process global events
Bad Mirror: AI tools scrape entire universe of art to generate on-demand clipart for Canva.com, now a $40bn company

Good Mirror: Someone sells a tool for AI-assisted drawing; something that actually helps people learn something, instead of doing the job for them, mostly by creating a bad photocopy

(comment deleted)
AI diagnosing disease better than any one doctor can. If we could build a model with a normalized dataset of everyone's medical history, we could catch a lot of diseases early. Especially if we encourage people to report even minor symptoms when they have it.

Imagine for example that everyone who gets diagnosed with pancreatic cancer reported repeated congestion of the left nostril three years before diagnosis. It's unlikely any doctor would ever make that connection, but an AI could, and then suggest regular pancreatic cancer screening to people reporting frequent left nostril congestion.

Agreed overall, but my concern about this is further alienating people on the margins with symptoms that don't fit into accepted disease frameworks.
There are many problems with the plan, and lots of guardrails would need to be added, but this was supposed to be a hopeful post of a future in which those problems are solved. :)

But you bring up a generally good point. Since AI magnifies what is already happening, at least in the short term we should expect it to increase marginalization of those outside the mainstream.

Lots of comments think health could be transformed for the better.

I wish AI would help detect, eradicate and replace toxic artificial materials around us. It only impacts health in the second-order or more. And they would need to be significantly incorporated down the supply-chain. But it is necessary.

Next 50-100 years:

Fully Homomorphic encryption if improved could allow for private cloud computing or really neat p2p networks.

Many diseases will become curable and some may be eradicated from the developed world or entirely

Next 100+ years:

Quantum key distribution could allow for the best possible encryption

Neutrino based communication could allow for wireless communication through the earth

Ongoing: Extreme poverty is falling world wide

AI will allow humanity to produce more intelligence with less (less labour) thus more spare-time for humankind. Like industrial revolution and mechanism did during the 19th century. Probably grey mirrorish though.
(comment deleted)
Something that could be very cool would be AI driven monitoring analysis and troubleshooting pre-work to support operative staff. Trying to figure out what's exactly going wrong in a complex, distributed system requires the system and the user to ingest a lot of information at times in order to make a guess what's going on. This is something an AI could really help with - ingest all of the information over some time period, and output a number of guesses or possible problems with probabilities and possible root causes assigned to them.

Think of an operator logging into the system at 3am with a C&C EVA voice calling and greeting them, "Hello Operator. We have 24 customer-facing systems offline. Preliminary analysis indicates that these are caused by 3 internal services and 2 postgres clusters being offline. With a confidence of 83%, the internal services are also offline due to the postgres clusters. I will now send you what I have found out about the postgres clusters with a probability of the scenarios and monitoring indicators for the different scenarios. It looks like the clusters are failing to elect a leader after several network-caused timeline switches with a confidence of 72%. Highly matching runbooks exist for 4 out of 7 highest probability scenarios"

Well, maybe I'd prefer that in text on a website, but no need to be that serious right now.

It is possible to get monitoring of this quality even with existing tools and maybe some internal extensions of these tools, sure. But that's a barrel you can pour effort into and it has no bottom at all. And it gets harder and harder the more complex and the more dynamic your environment is. It'd be great to dump all of that into a black box, even if that black box just establishes a global timeline of events.