Weird Science: Robot Pain, Tetris & Trombones

Richard Bradbury: A pretty broad palette for a delve into the weirder side of Mattsplained today. Everything from vintage gaming achievements to AI ethics, healthcare and space exploration. Or as Matt calls it, tangential normality.

Matt Armitage:

• We like to keep things topical and current on Mattsplained.

• That’s why we moved from Friday to Thursday. To bring you just that bit closer to the beginning of the universe.

• So this week we start with a story from the cutting edge of gaming

• Tetris to be precise.

• Willis Gibson, a 13-year-old competitive Tetris player from Oklahoma, became the first ever player to beat Tetris.

• In a world where people buy intricate, beautiful, AI enhanced games that takes years to create.

• And complete them in a weekend. Tetris is a bit of a standout.

• Released in: 1985

• Up until now, no human player ever forced the game on to it kill screen.

• AI has but It was thought to be impossible for human players.

• Were you a Tetris fan?

Richard Bradbury: replies

Matt Armitage:

• Obviously the first consoles had already come out in that era.

• But it was still the time of machines in arcades and cafes and bars.

• Where the top score would be permanently displayed along with player’s initials – usually 3 letters.

• So there was this sense of competition – being a local hero – if it was your name on the scoreboard.

• Of course, kids. Meaning the name by the top score would often be something offensive or vulgar.

• Much to the anger of the owner of the café or wherever the machine was.

• Anyway – I digress. Willis.

• Did you realise there was a Tetris scene with competitive players?

Richard Bradbury: replies

Matt Armitage:

• This dovetails back with something we talked about in last week’s episode.

• Generations moving away from cutting edge technology.

• I mentioned we’ll do something on that in the future, so I won’t go into the various subcultures, like retro gaming, here.

• Willis’ used a newly developed technique called rolling to beat the game.

• From what I understand, this involves holding the Nintendo controller upside down.

• So, your fingers are pushing up rather than down and you can roll the back of the controller with your thumb.

• Allowing for much faster control.

• Personally, I think it’s pretty cool. Tetris is my favourite game, even though it’s been a long time since I played.

• And I’m terrible at it. Apparently, some newsreader reported on the story and added:

• He should spend more time outside.

• Ok, boomer.

• When he’s living in his esports millions mansion and you’re sitting forgotten and incontinent in a care home, with nothing but an aloe vera plant for company, you might recall those words.

Richard Bradbury: Why an aloe vera plant?

Matt Armitage:

• Well, this show is all about empowerment.

Richard Bradbury: ad lib [it is, I hadn’t noticed]

Matt Armitage:

• Sorry, I mean alternative power.

• I always get those things confused.

• Aloe vera plants because in the future we may be using aloe vera plants to power health or other personal devices.

• Now I know how this sounds. Like I’ve drunk some hippy koolaid.

• I keep banging on about energy sources and battery technology.

• And one of the reasons is because people are coming up with incredible, brain-melting alternatives.

• And in this case it’s supercapacitors made from aloe vera plants.

• I found this on new scientist – of course.

• This is a research project from Beijing Institute of Technology.

• They’ve blended biology with technology to create a supercapacitor almost entirely from different parts of an aloe vera plant.

Richard Bradbury: Just for argument’s sake, can you explain what a supercapacitor is?

Matt Armitage:

• It’s like a battery in a lot of ways.

• It’s a device that stores energy and can discharge that power much faster than a battery.

• making them ideal for applications where quick bursts of power are needed.

• And it turns out that aloe vera plants – we’ve used the sap in their leaves for lots of purposes,

• Ranging from skin treatments to health food and drinks.

• The team used the outer layer of the aloe leaf which was heated to produce activated carbon for the electrodes.

• The juice inside the aloe was frozen to create an aerogel, acting as the supercapacitor's electrolyte.

• And the resulting capacitor, which measure an almost imperceptible 4 millimeters across,

• Can then be installed in living plants, was installed in living plants, including aloe itself.

Richard Bradbury: So, essentially creating e-plants?

Matt Armitage:

• The team is calling them electric plants and e-plants.

• Which I love the idea of. Imagine carry an iphone and an e-plant with you wherever you go.

• At the moment the energy storage is quite low, enough for small lights or to charge low-power devices like wearables.

• Zhao's team even envisions this technology as a power source for travelers in remote areas.

• My wife is an avid hiker. And I can see the appeal of taking an organic, biodegradable power source on a trip.

• The team is looking at the scale of the power generation.

• Making the cell larger to store more energy would mean a larger incision in the host plant, risking infection and damage.

set of possibilities to start: 2024

• Plant power.

Richard Bradbury: But what happens when AI realizes it can bypass the humans in pods and go straight to plants for power?

Matt Armitage:

• Yean, conspiracy nuts will have to introduce a green pill alongside their blue and red ones.

• I assume you’re talking about the story about there being a 5% chance that AI will destroy humanity.

• Lots of versions of this one floating around. This one is from NS.

a survey was conducted among: 2700

• Again – this is reported via NS.

• They were asked about their views on future AI technological milestones and the societal consequences.

• 5% is small but it’s one of those statistics that niggles away at you.

• One in twenty is something you might bet on as a long shot.

• By the way, 5% sounds a whole lot worse when you say one in 20.

• One of the survey’s authors, Katja Grace, from the Machine Intelligence Research Institute ,

• Told NS that it points to a general belief among AI researchers that there is a non-minuscule risk of advanced AI posing a threat to weaklings in meat suits.

• I mean humanity.

Richard Bradbury: So it’s not really about the percentage, so much as the perception that there is a – perhaps serious – risk there?

Matt Armitage:

• Part of it is down to that perception of uncertainty surrounding the technology.

• What is it, how does it work, how quickly will it evolve?

• And the more it evolves, the more opaque the answers seem to become.

• The greater the power – it’s not really power in human terms, it’s compute ppwer.

• So I guess it’s more correct to say the greater the ability of the technology, the more perceptually remote it becomes and the more uncertain we feel.

• And as as Émile Torres from Case Western Reserve University points out,

• AI researchers aren't necessarily the best at forecasting the future trajectory of AI.

• It's not about worrying unnecessarily but being prepared and aware.

• There are immediate concerns with AI, like deepfakes, manipulation of public opinion, and worsening economic inequality.

• These are issues we're already facing.

• So, I think we have to approach it with a mixture of caution and optimism.

• To be aware of that potential for risk and work towards some level of AI development that wider society views as safe and ethical AI.

• Rather than having that process dictated to us by the companies that profit from that rapid and unregulated development.

Richard Bradbury: Play us into the break with something weirder than weird?

Matt Armitage:

• Sticking loosely with the subject of extinction.

• I am available for party bookings and motivational speaking, by the way.

• We spoke about grief tech and services to preserve digital memories last year.

• And I think we also talked about AI that can take on elements of someone’s personality and talk to grieving friends and relatives.

• And even create limited video clones of a person.

• Psychologists seemed to be divided about its benefits..

• But one thing we weren’t sure about was people’s attitudes to these developments.

• This story is via iflscience

• A recent study by Dr. Masaki Iwasaki from Seoul National University School of Law explored the subject.

• It surveyed 222 US adults and found stark differences in opinion depending on certain factors.

• For instance, 97% of respondents felt it was inappropriate to digitally resurrect someone who was known to have disagreed with the idea,

• while 58% were okay with it if the person had given consent.

Richard Bradbury: I think that was one of the issues we raised during the previous show. The importance of consent, and the idea that there could be multiple copies of a person after death, as people create their own versions of the deceased.

Matt Armitage:

• Yes, as we mentioned at the time.

• The idea of recreating someone digitally raises ethical questions about identity and consent.

• It's not just about the technical feasibility but also about respecting individual autonomy, even after death.

• It also suggests that this is an area where people’s ideas are still forming as they open themselves to the idea of the technology.

• For example, 59% of respondents disagreed with the idea of their own digital cloning after death.

• Which is interesting. 97% feel it’s ok to clone someone else, but the idea of being clones gives the majority of people the ick.

• That’s an interesting discrepancy.

• Especially given that a new Elvis hologram show debuted at the start of the year.

• The dead are increasingly taking up space in the present.

• The idea of posthumous digital rights isn’t just going to be for famous people.

• It’s something that will increasingly occupy a space in everyone’s life and post-life if there is such a thing.

• In fact, Dr Iwasaki suggests that more people start to include clauses about digital cloning and how they are represented after death.

Richard Bradbury: Are you worried about how you might be represented after you die?

Matt Armitage:

• I’m a bit of a fundamentalist about this.

• Once I’m gone, I’m gone. I won’t care what the future decides to do with me.

• Unless I’m a futurama head in a jar.

• For me it’s always about the living. They are the ones who will be affected by any representations of me.

• Certainly, in terms of commercial exploitation – any proceeds should go to whoever I have designated.

• Third parties shouldn’t be able to take my voice, or image or words and use them commercially.

• Unless I’ve granted those rights, or my estate holders have granted those rights.

• I don’t like the idea of being turned into a tikitok dance meme after I die, but that’s living me talking.

• My dead self won’t care. So have at it.

Richard Bradbury: Are we heading to happier places after the break?

Matt Armitage:

• If you call artificial pain sensors happier, then sure.

Richard Bradbury: outro

BREAK

Richard Bradbury: So far we’ve learned more about such topical subjects as Tetris and Matt’s estate planning. If those weren’t enough of a trigger, apparently we’re going to talk about pain sensors.

Matt Armitage:

• Yes. So, just to clarify. This is a story about robots. From NS again.

• And it’s linked to AI and autonomy.

• We’ve used the example of smart factories where a floor of industrial robots has to be shut down before people can safely walk around.

• Not because the machines have any homicidal intent.

• Precisely the opposite – they have no notion of the pain and harm they can inflict on us.

• Because they’re machines.

• At the same time, advances in AI are hugely increasing what robots can do.

• We can now give them much more individual autonomy.

• They can perform more delicate tasks and adapt to changing circumstances.

• We’ve also had some success in creating sensors for them that mimic sensations of touch or smell.

• But those systems are still very limited.

Richard Bradbury: Like the hands that can pick up eggs?

Matt Armitage:

• Yes. So you could program a robot hand to pick up an egg.

• But until you changed the programming, all the objects would have to be eggs.

• But if the previous object was a brick, and the programming isn’t changed,

• Then most robots would break the egg. As they would exert the same force as they did with the brick.

Richard Bradbury: I like that your examples are either a brick or an egg.

Matt Armitage:

• I hate eggs – this isn’t a joke.

• They make me nauseous.

• So I think my subconscious goes to brick when I see someone eating one.

• Back to the subject of motor control – in human beings, it isn’t just about sensory feedback.

• It’s also about the sensation of pain and the anticipation of pain.

• Knowing that a pan is of oil is hot for example and to keep fingers out of it.

• Having similar systems in machines would enable them to act with more autonomy and mitigate risks to themselves and machines and people around them.

• To that end researchers at Hunan University in China, have developed an artificial pain-sensing system for robots.

• This is basically a layer of artificial skin that can sense external forces that might be harmful.

• In humans the process relies on nociceptors – pain sensors in our skin that send electrical signals to our brain.

• What the Hainan team Tan has done is replicate this with a crystal made of zinc and gallium.

• When a strong force is applied, it releases electrons, creating an electrical signal, much like our nociceptors.

Richard Bradbury: Meaning that the robot feels pain – or at least a simulation of pain?

Matt Armitage:

• Yes. So there’s. visual component too.

• Our bodies don’t just tell us there’s pain, they tell us where the pain is.

• To give the robots that same ability, the force applied to the crystal generates light flashes, which a camera monitors.

• This dual approach – both electrical and optical – allows the robot to determine the intensity and location of the pain.

Richard Bradbury: There’s a difference between force and harm. How do they differentiate between, say, a gentle touch and something that's actually harmful?

Matt Armitage:

• That's where the AI comes in.

• The algorithm with different electrical and optical readings from various objects - like a knife, a rod, and a cotton ball.

• This training helps the AI distinguish between harmful and non-harmful stimuli.

• I guess you could think of it as machine aversion therapy.

• Teaching it what it should be “afraid of”

• So far the results have been really positive.

• A robotic hand equipped with this skin and algorithm could differentiate between safe and harmful objects with a 97.5% success rate.

• This means it knows to hold something soft, like your brain, but drop something potentially damaging, like a cactus.

Richard Bradbury: Regular listeners will know that Matt introduces unpleasant ideas slowly. He mentioned a machine squeezing my brain, which leads me to think there might be medical or even surgical uses for this technology.

Matt Armitage:

• Bah. Foiled.

• Brain and other micro-surgery are already conducted with robots.

• Because you need tools that can work at a much smaller scale than a human hand.

• Sometimes even remotely.

• Obviously, these tools are controlled by a human surgeon.

• But there’s no feedback mechanism. It relies on the surgeon to guide the machine successfully and accurately.

• The researchers have tested the machines doing test biopsies.

• And the artificial skin on the bots prevented damage to the organs.

• Opening up possibilities for surgical robots, either autonomous or guided.

• Allowing a human surgeon to guide the machine remotely and still receive sensory feedback.

• Which would be an enormous step forward – enabling hard-pressed medical services to meet supply and demand with local theatres and remote surgeons.

Richard Bradbury: You want to stick with sensory organs for the next story, is that right?

Matt Armitage:

• Nearly. Another NS story.

• For decades it’s been suspected that human cells oscillate at certain frequencies.

• The example NS uses is a tuning fork. You bang the tuning and it vibrates and emits the hum of whatever frequency it’s tuned to.

• Things vibrate different amounts at different frequencies.

• The classic example is buildings and earthquakes and those conditions where lesser quakes could be more damaging to buildings

• because they start to oscillate against the quake vibrations rather than with them.

• So, obviously, at some frequencies, objects vibrate just a tiny bit and stop quickly.

• At others, resonant frequencies they move a lot more.

Richard Bradbury: And determining these resonant frequencies in human cells could be used as a diagnostic tool?

Matt Armitage:

• That’s the supposition, which I’ll get to in a bit.

• A team at the Spanish National Research Council has figured out the resonant frequencies for breast cells.

• The researchers created a tiny cantilever made of gold and siliconwhich they used to pick up individual human breast cells from a dish.

• Because the cantilever was so small it vibrated in response to the cell’s own motion.

• By illuminating the cantilever with a laser, the researchers could then measure this subtle motion from the way the light reflected.

• And calculate the resonant frequency.

• Interestingly, the insights leading to the breakthrough came when the team was investigating other aspects of cell behaviour.

• And they noticed that the cells were vibrating.

• They hadn’t imagined that human cells would behave like a tuning fork.

• As they admit themselves, if you’d hypothesised that human cells behave like a tuning it would sound far-fetched.

• So seeing it was a surprise.

Richard Bradbury: Do the cells vibrate at an audible pitch?

Matt Armitage:

• That’s the question, isn’t it.

• It gives new meaning to that idea of listening to your own body.

• So it turns out that one of the resonant frequencies of the cells fell between 150 and 180hz, which is outside the realm of human hearing.

• However, they found it also resonated at between 10 and 30 khz,

• Which is around where the audible spectrum turns into ultrasound.

• Which I guess means whales may be able to hear our cells but we can’t.

• What does any of this mean?

Richard Bradbury: Yes, what does any of this mean?

Matt Armitage:

• I’m glad you asked me that.

• As you mentioned earlier – it opens the door to diagnostic research.

• If we can come up with values for a variety of normal cells, we can use those resonant frequencies to determine if the cells are healthy.

• Or if they have mutated or are under attack.

• Obviously, it would be some kind of targeted sound device, but I’ve got a mental picture of making someone stand on one of those vibrating exercise pads,

• The ones they tempt you to try at those places that sell massage chairs.

• Shaking you like a milkshake and then printing out the results of your full body scan.

• But there is also the possibility of using sound waves as a therapeutic tool.

• To target and destroy unhealthy cells, perhaps as an alternative to treatments like chemotherapy.

Richard Bradbury: Another quick and weird one to play us out?

Matt Armitage:

• Yes, sticking with sound waves and the idea of playing out.

• Scientists have discovered radio waves in space that demonstrate a diminishing frequency.

• Which is something that they’ve never seen – or rather heard – before.

• That they’re likening to a sad trombone effect. Everyone knows what that is right.

• Play Clip: https://www.youtube.com/watch?v=CQeezCdF4mk

• Can we play that again?

• Play Clip: https://www.youtube.com/watch?v=CQeezCdF4mk

• No reason for that second one, just a gratuitous wah wah wah.

radio burst (FRB) called FRB: 202209

• over the course of 541 hours of observation with the Allen Telescope Array in California.

• Which consisted of 35 of these waves with a diminishing tone.

• So it’s like each of these waves is part of a really long sad trombone effect.

Richard Bradbury: Is it evidence of intergalactic vaudeville comedy?

Matt Armitage:

• If only that were true. Douglas Adams was a biographer rather than a novelist.

• The SETI team seems to think that they are most likely to come from magnetars.

• Spinning, highly magnetised neutron stars.

• Which are the likely candidate for some of these fast radio bursts.

• But, as they haven’t seen anything like, they aren’t sure which sub-category of magnetars could be responsible.

• Play Clip: https://www.youtube.com/watch?v=CQeezCdF4mk