Suddenly, paperclips. Billions of them.

So I may have spend some of my idle brain and CPU time yesterday making paperclips. Wait! I promise this is more interesting than it sounds!

The game Universal Paperclips, shown to me by a friend (thanks Wil!) after my Dangerous AI article, sees you take on the role of an AI whose existence has one simple goal. Make paperclips.

It starts off as you might expect. You have to balance costs and earnings, create better manufacturing equipment, improve efficiency and so on. Up until the point where you take over the world, learn how to process any kind of matter into paperclips, learn how to make machinery out of the paperclips themselves, an ultimately dismantle the entire planet and turn it into paperclips. Then, once you’ve finished demolishing the Earth, your raison d’être continues by exploring space with the ultimate goal of turning all matter in the universe into paperclips. Obviously.

While not a game to focus all your attention on, it’s amusing to have running in a background window. As I write this, my self-replicating probes have explored 0.000053646601% of the Universe and are creating 258.9 quattrodecillion paperclips per second. Feel free to speculate on how they’re stored in order to prevent them from spontaneously collapsing into black holes due to their high iron content. I’m not sure quite what makes this game so addictive when, involving little more than numbers and idle clicking, it should be tedious by all rights. Perhaps it just amuses my inner evil scientist.

Interestingly, there’s a serious background to this odd little game. It’s based on the Paperclip Maximizer a thought experiment by Nick Bostrom, a Swedish philosopher, dating back to 2003. The point to it is illustrating the potential existential risk which could result from an artificial intelligence with a seemingly innocuous goal. While it may have no inherent malice in its intentions, its actions could pose a threat if allowed to spiral out of control. In this case, if its sole goal is to create as many paperclips as efficiently as possible, it could be justified in its decision that humans are unnecessary. Likewise planets, stars, galaxies…

In he 2008 book Global Catastrophic Risks, AI researcher Eliezer Yudkowsky wrote a chapter entitled Artificial Intelligence as a Positive and Negative Factor in Global Risk. He summed up the concern succinctly:

❝ The AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else.❞

Eliezer Yudkowsky

The idea is, of course, hyperbole. Bostrom didn’t genuinely believe that a rogue AI might turn the Universe into paperclips (hilarious and horrifying though that might be), rather this is an example of a hypothetical phenomenon dubbed Instrumental Convergence.

That term may seem a little opaque, but it’s fairly simple really. Basically, any intelligent agent, human or otherwise will have a set of goals when attempting to accomplish something. A final goal, containing the intended end state, and a set of intrumental goals which are required to get to that end state. For example, I currently have a final goal of writing a blog post, which requires several instrumental goals such as organising my thoughts, typing sentences, structuring them into paragraphs, creating a logical narrative, etc.

❝ Several instrumental values can be identified which are convergent in the sense that their attainment would increase the chances of the agent’s goal being realized for a wide range of final goals and a wide range of situations, implying that these instrumental values are likely to be pursued by a broad spectrum of situated intelligent agents.❞

Nick Bostrom

Going back to the hypothesis, instrumental convergence is the idea that whatever an intelligent agent may set out to do, its most basic instrumental goals may tend to be very similar. At a fundamental level, you could consider them to be analogous to the instincts found in animals. All animals have a set of basic instrumental goals, like avoiding danger, eating enough food to survive, reproducing, continuing to breathe, and so on. For an AI, we’re off the edge of the map, but a few basic AI drives have been proposed, including utility function, self-improvement, freedom from interference, unbounded acquisition of resources, and self-preservation.

Linking back to my Astrotropes post, this idea has been explored in fiction too. The 2014 movie Transcendence features an AI which takes on all of these basic values. The result is unsettling. So could a situation like this actually occur? The answer is, we genuinely don’t know.

I’d argue that two of those proposed AI drives are actually quite human in origin and may not necessarily apply to an artificial intelligence.

When you think about it, unbounded acquisition of resources isn’t logically necessary for all purposes, and even where it is a goal, it may not necessarily mean the same as acquire all available resources by any means. I’d suspect that living in a capitalist society where acquisition of material wealth is the main driving factor might skew our perspective on this. If your goal is to manufacture as many paperclips as possible, then acquiring as much matter as possible is a logical solution. For most other purposes, it seems more likely that an intelligent agent may wish to acquire precisely as many resources as necessary and no more.

Freedom from interference too, sounds like the kind of thing a comic book villain might request, but it seems unlikely to be something an AI might require unless it was proven to be a necessity. And even then, particularly if utility function is a basic AI drive, it seems unlikely that an AI would enact any dangerous or apocalyptic feats to accomplish it. That would be an unnecessary and illogical waste of effort and resources which could be better used elsewhere. To quote GLaDOS in Portal 2, “The best solution to a problem is usually the easiest one. And I’ll be honest, killing you is hard.”

Some people take the ideas of instrumental convergence to mean that we should try to make certain that an AI has implicitly human values to prevent it from doing harm. I’d add that giving it a purpose which is more nuanced than simply “do as much of this as you possibly can” would probably be a good idea too. As any good coder will tell you, infinite loops are best avoided. In any case, thought experiments and instrumental convergence aside, I’m still not convinced that artificial intelligence is the great existential threat that so many people seem to think it is.

At least, unless paperclips are involved. Then we’re clearly doomed.

Advertisements
Posted in Sci Fi, Tech | Tagged , | 1 Comment

Astrotropes: Dangerous AI

This trope may not be particularly astro-related, but it’s nonetheless quite ubiquitous in science fiction, both in space and on Earth. For a long time, humans have daydreamed about the ability to create artificial intelligence (AI) which may be equal to us, or even surpass us. But as technology increases to the level where it’s starting to seem like a possibility, it seems our paranoia about the idea is steadily increasing.

Particularly in recent years, a lot of fiction focussing on AI seems to be full of  themes centred on moral dangers, othering, and generally portraying AIs as being a danger to us. This is not being helped by influential figures like Stephen Hawking and Elon Musk stoking our paranoia in the real world.

This article discusses, and contains some pretty major spoilers for, Star Trek: The Next Generation (Episode 1:13, Datalore), Prometheus, Alien: Covenant, Ex Machina, Portal, 2001: A Space Odyssey, the Terminator movies, the Matrix movies, and Frankenstein.

Continue reading

Posted in astrotropes, Sci Fi | Tagged , | 2 Comments

Energy!

Inspired by that last Astrotropes post of mine, here is a very circular graphic showing kinetic energies of various things, fictional and non-fictional. Go and check the other post for some more details, and mind out for any spoilers.

(click to humongify)

Essentially, there are only two types of energy in the universe. Kinetic energy and potential energy. Moving objects (like asteroids and tennis balls) have kinetic energy, and if a moving object strikes something, that energy is transferred. Pretty much everything also contains potential energy, waiting to be released. That could happen by detonating it, in the case of a stick of dynamite, or by eating it in the case of a doughnut. Also, a doughnut contains more potential energy than a stick of dynamite! How wild is that!

The sizes in the graphic are scaled logarithmically. Double the size means 10 billion (1010) times as much energy. Which makes some things a little tricky to discern, but it was the only realistic way to fit a black hole collision and a cosmic ray particle on the same graphic. I mean, I tried a linear scale, but, well…

So yeah, a black hole collision releases about a hundred thousand million trillion trillion trillion⭐️ times as much energy as a cosmic ray particle, and I figure most people don’t have screens that large.

Also, yes, the most energetic cosmic ray particle ever detected was nicknamed the OMG particle. With good reason too. A single proton travelling at 99.99999999999999999999951% the speed of light, and consequently carrying the same energy as a baseball is not the kind of thing we normally detect here on Earth!

⭐️ This sounds like hyperbole, but it’s really not. That particle was travelling with 48.07 J of energy, while two colliding black holes release 5.39 × 1047 J!

Posted in physics, Sci Fi | 1 Comment

Astrotropes: Spaceship collisions

High above Earth, several United Federation starships are locked into a desperate battle against a Borg cube. On the bridge of his ship, Worf hears a grim damage report. His blood boils and he slams a console with furious fists. “Perhaps today is a good day to die,” he growls, before barking an order to his surviving crew. “Prepare for ramming speed!”

This post is loaded with massive (and high velocity) spoilers for Star Wars – Rogue One, A New Hope, Return of the Jedi, and The Last Jedi.

The above sequence with Worf fighting against the Borg, by the way, is from Star Trek First Contact. It’s not a spoiler because it happens in the first 5 minutes of the movie!

Continue reading

Posted in astrotropes, Sci Fi | 3 Comments

Astrotropes: Surviving Space

Space, as a background, is a mainstay of sci fi. But space is hazarous, and if you’re to find yourself without air to breathe, that should probably be cause for alarm. Naturally, sci fi characters (much like real life astronauts) go to great lengths to avoid this. “Don’t leave your spacecraft without a protective suit” is generally some rather sound advice. Unfortunately, in many stories, that isn’t always an option.

Caution: This post has mild spoilers for Star Wars VIII: The Last Jedi, Sunshine, Gravity, and episode 7 of Cowboy Bebop.

Continue reading

Posted in astrotropes, Sci Fi, space | 2 Comments

“We humans are capable of greatness”

Happy Birthday Carl ❤️

Posted in unassigned | Leave a comment

Quark Fusion

When thinking about the future of energy production, aside from renewable energy like solar power, fusion is something which immediately springs to mind. Clean, efficient, and the same thing which stars have been using for around 13 billion years. But imagine for a moment, if there was another energy source available which was dramatically more potent.

A paper published recently in Nature discusses the idea of quark fusion. For anyone who’s uncertain what that means, it’ll require a brief trip down the rabbit hole.. So matter, as you probably know, is made up of atoms. And atoms are made up of electrons and nucleons. Nucleons come in two types, protons and neutrons, and are a subtypes of a whole family of subatomic particles called baryons. Baryons, in turn, are made up of quarks. Quarks, as far as we can currently prove, are the most fundamental building block of matter there is, and they come in six types. Every baryon contains three of them – so, as an example, in its most fundamental terms, a hydrogen atom is a structure made from two up quarks, one down quark, and an electron.

Still with me? Cool. So Marek Karliner and Jonathan Rosner, a pair of physicists working in Tel Aviv and Chicago respectively, published their work all about the idea of directly fusing quarks. Specifically bottom quarks.

Now, I’m not a particle physicist or a nuclear physicist, so I may be missing a few of the finer points in all of this, but the concept concerns bottom quarks – the second heaviest type. One of these quarks alone is already over 4 times as massive as an entire proton. Apparently, fusing two of them releases a hilarious amount of energy.

How much energy? Here’s a comparison.

238U ⟶ 92Kr + 141Ba + 3 1n ……………… ~0.9 MeV

3He + 3He ⟶ 4He + 2 1H ……………….. 12.86 MeV

Λb + Λb ⟶ Ξbb + 1n ………………………… 138 MeV

Splitting uranium into krypton and barium produces about 0.9 MeV of energy per reaction and spits out 3 neutrons (most fission reactions give out roughly this amount of energy). Fusion of helium-3 is the most energy efficient fusion reaction, producing one helium-4 atom, and 2 protons, and gives 12.86 MeV of energy.

The symbols in the third reaction may be less familiar. Λ and Ξ are types of baryon (quarks don’t really exist outside baryons or mesons), but the important part is those b symbols. Λb is a lamba baryon containing one bottom quark. Ξbb is a xi baryon containing two bottom quarks. Karliner and Rosner describe this as being a quark level analogue of a nuclear fusion reaction.

The amazing part is the energy released! The quark fusion reaction releases 138 MeV, which is 10.7 times as much as helium-3 fusion, and 153.3 times as much energy as a nuclear fission reaction!

The biggest difference is that the other reactions here are chain reactions. The additional products can go on to set off additional reactions, which is what makes them useful in energy production (and scary in weaponry). Just like Meitner, Karliner was concerned by the potential implications of such a huge energy release from a single reaction. However, quark fusion is exactly that – a single reaction. Bottom quarks only exist for a fleeting moment at most, so with no means of easily producing or storing them, and no means for a chain reaction, a quark fusion reaction like this is just a one off.

So sadly, there’s not much scope for any future source of energy here. At least not yet. Guess we’ll have to stick with solar and trying to make helium-3 fusion economically viable.

Karliner & Rosner (2017)

String theorists will argue this point, and as soon as someone has evidence that strings exist, I’ll gladly concede.

Posted in physics | Tagged | 2 Comments

Aurora on your Finger

I’ve been very fond of jewellery pretty much forever, so I have a pretty sizeable collection of rings by now – but this aurora borealis ring made by Secret Wood is on a whole different level.

Isn’t that gorgeous?

Secret Wood make these beautiful jewellery pieces entirely from wood and jewellery resin, with a few select pieces including other things, like flowers, trapped inside the resin.

These rings made to order and, as you might expect, each one is unique due to the natural variations in the wood and materials they’re made from. But to be certain you’re happy with your purchase, they’ll send you a photograph if your ring before shipping it. They’ll also cost you somewhere around $110-$170 US. For jewellery like this, that seems like a very good price, to me!

Their website has a whole collection of different ring styles (including limited editions) which, quite honestly, seem like the sort of thing a wizard might wear. Or if you want even more pretty things to look at, they have an instagram page too.

Posted in art | Tagged | Leave a comment

Space Rocks

I stumbled across this rather lovely size comparison of small Solar System objects, and I thought I’d share. Click it for a larger view!

Taking pride of place is Ceres, empress of asteroids, with her smaller sibling Vesta, the largest asteroid in the main belt. To the left are a handful of more typical asteroids, all of which have been seen up close by robots we’ve sent out to investigate them.

Below are 3 of the Solar System’s huge collection of moons, and the ever curious asteroid Chariklo with its ring system. At the very top, near the right, are the two Martian moons, Phobos and Deimos which really does illustrate how puny they are, even compared to some asteroids.

This image was created by Daniele Bianchino, left in a comment on the Dawn Blog a couple of years ago. His website’s quite a visual treat, even if you don’t speak Italian. Nice work, Daniele!

Posted in astronomy, space | Tagged , | Leave a comment

An Uncomfortable Truth About Spacecraft

Since Juno’s arrival at Jupiter, a lot of people noted how impressively technology has improved, with Juno being powered by solar panels. Indeed, without the progress we’ve made, Juno wouldn’t have been possible. It may not be immediately apparent to everyone, seeing as solar panels are so synonymous with satellites, but they’re one of the most distinctive things about Juno. No craft has ever been sent to the outer solar system before with solar panels as their energy source.

Solar power is an abundant source of energy on Earth, but a spacecraft at Jupiter receives only about 4% as much solar energy as one in Earth orbit does. Juno’s solar panels are huge, and would harvest 12-14 kW of power in Earth orbit. Out by Jupiter though, they only produce a meagre 486 W. Which is why all previous spacecraft were actually powered by nuclear isotopes. Sadly, this fact makes the future of spacecraft visiting the outer solar system rather uncertain.

Vehicles which can’t be practically powered by solar energy, like New Horizons, or vehicles which need a lot of power, like Curiosity, are traditionally powered by radioisotope thermal generators, or RTGs for short. The type used by Cassini and New Horizons looks a bit like this:

They’re quite ingenious devices really. An RTG is a generator with no moving parts at all. Instead, it’s filled with a well chosen radioisotope. As that isotope decays, it produces heat, and an array of thermocouples surrounding the hot material convert that heat directly into electricity using the Seebeck effect. These are such a convenient way of generating a sustained amount of power that the Apollo missions used several of them, and the Soviet Union even found uses for them here on Earth. Sounds pretty useful, right? The trouble is that there are only a very few radioisotopes which work effectively in an RTG. The one traditionally used produces enough power that it can easily glow red hot using only its own internal heat. Plutonium-238.

Plutonium is an almost entirely human-made element. It’s technically the heaviest naturally occurring element, but only ever formed in trace amounts as uranium-238 naturally captures stray neutrons.  And, quite rightfully, plutonium has a bad reputation. While Dummkopf Tinyhands may like to bemoan the general badness of uranium, plutonium is much more scary. Back during the anxiety-inducing days of the Cold War, when everyone was inexplicably obsessed with nuclear weapons, one thing which nuclear reactors were producing was Plutonium-239 (Pu-239). This is the frightening weapons grade material which… well. This is a subject which keeps me up at night, which is best left for another time. The thing is, nuclear reactions are a bit like chemical reactions in that they almost never produce just one product. Most nuclear reactions tend to give byproducts.

One byproduct of the reactions to make Pu-239 is Neptunium-237 (Np-237), and if you irradiate Np-237 you can produce the very useful Plutonium-238 (Pu-238). In fact, a lot of the Pu-238 produced for power sources also contained a fair percentage of Pu-239. In other words, the power source for all the deep space exploration NASA has ever done was actually a direct consequence of the US and Soviet military producing nuclear weapons. Which is a sobering thought.

Plutonium can keep providing a spacecraft with power for a long time♣︎. Voyager 1 for instance, now classified as an interstellar mission, was launched in 1977 and is not expected to finally run out of power until 2025. It left carrying 37.7 kg of Plutonium (a schematic for Voyager’s RTG is shown to the right here) which gave it an initial power output of 470 W. As of 2015, that had dropped to about 255 W.

Mercifully, the world is producing less Pu-239 now. For decades, humans realised that having enough nuclear weaponry to extinguish all complex life on Earth’s surface and phrases like “mutually assured destruction” were probably bad. The thing is, just like the Plutonium being carried by Voyager 1, the Plutonium that stayed behind is also decaying and is gradually producing less and less useable power for RTGs. In short, as a species we’re running out of Plutonium.

NASA, of course, are well aware of this problem. In 2008, NASA Administrator Mike Griffin explained in no uncertain terms, “After [Curiosity] launches, we’re effectively out of plutonium.” Since then, dwindling reserves in the US have been supplemented by remaining reserves in Russia. Unfortunately, NASA only has about 35 kg of Plutonium left to work with. Only about 17 kg of that stockpile is still good enough to be used in RTGs, and that’s been reserved in its entirety for the next Mars rover mission, currently scheduled for 2020.

So what’s to be done?  In 2013, the US Department of Energy (DoE) reactor pool at Oakridge started making Plutonium again for the first time in decades, but it’s going to take the best part of a decade before enough Pu-238 is produced to actually be useable. The DoE stance is that they’re producing enough for currently projected NASA missions and that there’s no current shortfall, but this has been rightfully criticised as circular reasoning – of course NASA aren’t going to plan anything using energy sources they don’t currently have.

Meanwhile, the DoE are researching new production methods, presumably because they’re no longer tied to methods for making weapons grade materials. Or perhaps as a way of actively not making such materials. One method they’re researching involves using pellets of pure neptunium oxide.

Another option may be to use a different nuclide. Unfortunately, the requirements are quite specific. The nuclide used in an RTG needs to be have a long enough half life that it provides consistent amounts of energy for a suitable amount of time, while also decaying rapidly enough that it gives out a useful amount of energy. That energy also needs to be a useful type. Alpha emission is ideal, while beta and gamma ultimately require heavy shielding. For spacecraft, it also needs a good energy density (energy output per kg) given that when you’re trying to send something as far away as Jupiter and beyond, every last gram really does matter.

Sadly, out of the 905 known nuclides with half lives of over an hour, only about 30 are known to meet these requirements and be suitable for use as spacecraft fuel. Most of them decay much more rapidly that Pu-238 too, meaning that while they provide a good source of energy, they won’t be effective for the long timescales required for long distance space travel. Additionally, none of them are safe to use with as little shielding as Pu-238.

While it’s not impossible that we may still discover another good isotope for use in RTGs – or better, a different type of power source altogether – for now, it looks like Plutonium is the best we’ve got.

Looks like we’d better keep those reactors running a little while longer.

For anyone who knows a little about how solar panels work, the thermocouples in an RTG actually work in a very similar way. Essentially, you take two metals at different points in the galvanic series and join them at two places. The electron energy levels shift differently in the two metals, with the result being an electrical current. More sophisticated designs actually use n-doped and p-doped semiconductors, making them even more similar to solar cells.

♣︎ As you might expect from such a massive element, Pu-238 is a prolific alpha emitter. The heat it generates comes from its rapid decay into U-234, and with the alpha decay being the dominant mechanism, there are relatively few complications like chain reactions and unwanted byproducts. Alpha emission is also not very penetrating, so Pu-238 powered RTGs don’t require as much shielding as some other isotopes might.

Flying long haul to Neptune is a killer.

Images:
Schematic of one of NASA’s general purpose RTG units.
One of the plutonium pellets used in Cassini’s RTG.
Older design of RTG used in the Voyager probes.
DoE image of a 250g sphere of plutonium-238 (IV) oxide.
New Horizons before launch, with its RTG on display.

Posted in chemistry, physics, Tech | Tagged , , | Leave a comment