Host: Nick Petrić Howe
Welcome back to the Nature Podcast. On this week’s show, the computer that can debate humans. I’m Nick Petrić Howe.
Host: Shamini Bundell
And I’m Shamini Bundell.
[Jingle]
Host: Shamini Bundell
First up on the show, it’s time for a bit of a debate.
Interviewer: Nick Petrić Howe
In 2019, a historic debate occurred.
Project Debater
Greetings Harish. I have heard you hold the world record in debate competition wins against humans, but I suspect you’ve never debated a machine. Welcome to the future.
Interviewer: Nick Petrić Howe
That right there is the voice of an artificial intelligence called Project Debater, who came up with the statement and the whole debate herself. In fact, she has successfully debated humans live several times. On the podcast, we’ve talked about a few different AIs – ones that can play board games, some that can play video games, and AIs like Watson who competed on Jeopardy. But for the principal investigator of Project Debater from IBM, Noam Slonim, these challenges aren’t tough enough.
Interviewee: Noam Slonim
Although it is clear that all these grand challenges were extremely instrumental to the development of artificial intelligence, these board games still lie in what we refer to as the ‘comfort zone’ of artificial intelligence.
Interviewer: Nick Petrić Howe
Things like board games and video games have a clear ‘win’ state that an AI can try hundreds of techniques to achieve. Not so in the case of debate. Here, winners are tricky to identify, but now IBM have developed such a system by honing a few different AI technologies, such as argument mining and understanding of human language, and getting each of these different AI components to work together all whilst squaring off against an opponent. Project Debater, who looks a little bit like the monolith from 2001: A Space Odyssey, has been able to face off against top-tier human debaters in front of a live audience, such as the debate I played earlier which took place in California against Harish Natarajan, a world-class debater. Despite this challenging environment, Project Debater has done pretty well for herself.
Interviewee: Noam Slonim
In the three public debates that we had, we lost one, and we were able to win another one and it was nearly a tie in the third one, and in addition I think it was interesting to note that in all the debates that we had we also asked the audience another question: which side better enriched your knowledge during the debate? And in all of the debates, Project Debater obtained clearly better scores than the human opponent on this question, which was perhaps expected but still nice to see.
Interviewer: Nick Petrić Howe
This week in Nature, Noam and his team are publishing a comprehensive rundown of the technology involved in Project Debater. But before we get into how the system works, it’s worth quickly running through how the debates it participated in were laid out. To start, there’s a motion – what are we arguing about here? For example, should pre-schools be subsidised? Next, there’s opening remarks – four minutes of argumentation about why your side is right. Then after hearing what the opponent has to say, there’s four minutes of rebuttal. And finally, two minutes of closing remarks. So, at the start, Project Debater is given a position to argue for and then has 15 minutes to prepare the opening remarks.
Project Debater
For decades, research has demonstrated that high-quality preschool is one of the best investments of public dollars, resulting in children who fare better on tests and have more successful lives than those without the same access.
Interviewer: Nick Petrić Howe
So, here are some of the remarks that Project Debater is coming up with. But how is she doing it? Well, she draws on a wealth of human argumentation.
Interviewee: Noam Slonim
First of all, it has a large collection of around 400 million user articles from LexisNexis, nearly 10 billion sentences. And when the debate starts, the system is using various AI components to detect short pieces of text that should satisfy three criteria. They should be relevant to the topic. They should be argumentative in nature – that is they should argue something about the topic, not just be relevant. And finally, they should support our side of the debate. Then after finding these short pieces of text, the system is using the other AI components, like text clustering and et cetera, in order to glue these short pieces together into a meaningful narrative.
Interviewer: Nick Petrić Howe
Now, finding these bits of text is a huge challenge. Identifying the bits that meet those criteria are part of what Noam’s team have been working on since the project started in 2012. Project Debater largely achieves this through having lots of data and ranking how relevant those bits of sentences are. She then groups these sentences into topics using Wikipedia to help. After all that, Project Debater has to assemble those sentences into a coherent argument. Again, not an easy task. It’s hard to say exactly what it is that makes a debate compelling. And then there’s rebuttal.
Project Debater
For starters, I sometimes listen to opponents and wonder: what do they want? Would they prefer poor people on their doorsteps begging for money?
Interviewee: Noam Slonim
You need somehow to respond to the arguments of the opposition, and this starts by really understanding the words articulated by the human debater, and for that purpose, we used Watson’s speech recognition capabilities out of the box, but of course you need to go beyond the box. You need somehow to understand the gist of the human speech and the main claims being raised, and to that end we developed several techniques that typically rely on the same principle of trying to anticipate in advance what kinds of arguments the opposition might use and then listen to determine whether indeed the opposition was making these claims and then respond accordingly.
Interviewer: Nick Petrić Howe
Altogether, as Noam mentioned, Project Debater was pretty good at debating. Compared with other AIs, Project Debater was ranked more highly by audiences and close to that of a human expert, but not without some limitations. Sometimes, she found it difficult to make an argument flow like a real human does and, not too surprisingly, she kind of argued like machine, with lots of facts and figures and not as much emotion as a human debater. But whether or not Project Debater is better or worse than a human, it’s a big step for the field. Finding arguments in human written text – so-called argument mining – and language generation, well, these are tricky tasks for an AI, as AI researcher Elena Cabrio explains.
Interviewee: Elena Cabrio
Even for humans, for which debating is among the primary cognitive activities, I mean, with every day, so even when we debate, we need to apply a wide range of language understanding and language generation capabilities. So, for a machine, being able actually to address all these tasks at the same time in an automated way is actually a big improvement in the field.
Interviewer: Nick Petrić Howe
For Elena, who was not involved in this project, the ability of AIs to search through vast amounts of text and fine-caught arguments could help with an ever-growing problem in the modern world – information overload.
Interviewee: Elena Cabrio
The growing of the web, the increasing number of texts and data that are published every day, have actually highlighted a need to process such data in an automated way, to be able to identify, structure and summarise this huge amount information. People like us are more and more exposed to information – online newspapers, blogs, online debate platforms, social networks – so argument mining has actually the potential to help us with that because it provides the techniques to sift through this ever increasing amount of data and provide us with the relevant evidence items that we can find in them.
Interviewer: Nick Petrić Howe
Noam feels similarly, and sees Project Debater as more of a collaborator with humans, potentially helping them find arguments to assist their own debates or speeches. For both Elena and Noam, the next steps for this kind of research are to try and improve these AIs abilities to understand language as humans use it. What makes a good argument? What is convincing? Why is this compelling? But for now, Project Debater has given an insight into what this sort of technology could look like.
Project Debater
Thanks for this final opportunity to speak out in this debate and thanks, Harish Natarajan. One might say that this conversation can serve no purpose anymore, but I feel differently.
Interviewer: Nick Petrić Howe
When you saw Project Debater up there debating with people, what did that feel like?
Interviewee: Noam Slonim
I was proud.
Interviewer: Nick Petrić Howe
That was Noam Slonim from IBM Research. You also heard from Elena Cabrio from Université Côte d’Azur in France. If you feel the matter is still up for debate then you can find a link to Noam’s paper in the show notes.
Host: Shamini Bundell
Right now, though, Dan Fox is here with this week’s Research Highlights.
[Jingle]
Dan Fox
When it comes to regenerative powers, I’m going to stick my neck out and say that sea slugs might have a head start over other animals because some are able to grow a whole new body from just their severed head. Researchers observed that two species of the sacoglossan sea slug sometimes severed their entire bodies from the neck down. Their heads then continued to move around and eat algae. A slug’s head regenerated a heart within a week and the entire body within three weeks. Slugs more than a year old did not have this ability. They simply died when their heads were severed. The disembodied heads can’t digest the algae they consume, and the researchers hypothesise that the heads rely on chloroplasts in the algae, structures that use photosynthesis to make energy that can power the slug until its digestive system regenerates. It’s still unclear why the slugs shed their entire bodies, but the team thinks it might be a way to rid themselves of parasites. Get your heard around that research at Current Biology.
[Jingle]
Dan Fox
Archaeologists have uncovered jewellery that suggests some ancient European civilisations may have held certain women in high esteem. A silver diadem, amongst other ancient jewellery, discovered in a tomb in ancient Spain suggests that the woman buried there some 3,700 years ago was very high-status, perhaps even a young queen. The tomb sits below a sprawling compound that researchers identify as a palace, the oldest known from western Europe’s Early Bronze Age. The woman was buried beside a male partner, but the researcher link most of the 29 objects interred at the site to her. The items include silver headdress ornaments and earlobe plugs, along with objects that signified social status, such as pots with intricate silver plating and daggers with silver-plated handles. Dig up the rest of that research at Antiquity.
Host: Nick Petrić Howe
Finally on the show, it’s time for the weekly Briefing chat, where we discuss a couple of articles that have been highlighted in the Nature Briefing. Shamini, what’s caught your eye this week?
Host: Shamini Bundell
This week, gravitational waves have caught my eye. Not literally, but it is about how people are detecting them and how we hope to detect even more in the future.
Host: Nick Petrić Howe
Okay, but yeah, I remember vaguely that gravitational waves were detected a few years ago but remind me, how exactly do you go about detecting them and how might these new detectors fit in to this?
Host: Shamini Bundell
Right, yeah, so you probably remember the name LIGO, so that’s this gravitational wave detector in the US that published some of the first big results. There’s also VIRGO in Italy, KAGRA in Japan, and they’re all basically these big L-shaped structures, and the idea is that gravitational waves, which come from these huge events like black hole mergers or neutron star mergers, they cause the fabric of spacetime to stretch. And what that means for your sort of L-shaped gravitational wave detector is that the whole structure literally stretches, and the aim of the two arms is to compare the two and see if one of them stretches more than the other based on how fast light takes to go down it. Now, when I say stretch, we are talking tiny, tiny amounts, so less than the width of a proton.
Host: Nick Petrić Howe
Right, so it has to be an incredibly sensitive device then to detect these tiny, I guess, ripples in spacetime, so what will these new proposals, what will these new systems do to become even more sensitive?
Host: Shamini Bundell
So, there are two proposed plans here for the sort of next generation of gravitational wave detectors, and two slightly different methods that are being used, so one from the US and one from researchers in Europe. So, the main aim here is we have been able to detect gravitational waves from massive events, so two black holes whirling in to each other, for example, or neutron stars spiralling together in a killer nova, but the hope is that we could build a detector that could detect much more sensitive readings and just way more of them. And, well, my favourite quote in this article is the explanation from someone on the US team about how they’re going to go about doing this and the quote is, ‘We’re just going to make it really, really big,’ which is perfect. And it makes sense because these detectors are already huge – the one in the US, LIGO, has two arms in this L shape and each arm is 4 kilometres long – and the plan is let’s make one with arms 40 kilometres long. Actually, the sort of most amusing problem that they mentioned they might have with building a gravitational wave detector with 40-kilometre-long arms is that the arms have to be perfectly straight because it’s all about light sort of bouncing up and down, but the Earth curves, so they might have to raise the ends up way high into the air in order to get it flat.
Host: Nick Petrić Howe
So, bigger is better in one way – it seems like a typically US solution to things – but what was the method proposed by the European scientists?
Host: Shamini Bundell
I was trying to avoid saying that. I didn’t want to offend our US listeners but, yes, it does seem to fit the stereotype. The European physicists – I’m not sure whether there’s a stereotype about this – but they’re going underground, building a subterranean gravitational wave observatory called the Einstein Telescope, or at least that’s the plan. Both of these projects haven’t quite got off the ground yet, but that’s the hope, and this would be only 10 kilometres long but being underground would help to shield out vibrations, and they’re using various other things to try and maximise their ability to detect in particular very low-frequency gravitational waves.
Host: Nick Petrić Howe
Oh, well, hopefully in the future then we’ll get even more gravitational wave detections. And, well, keeping in a sort of space theme, the story I was looking at this week is about how 2020 was a record-breaking year for the number of asteroids near Earth.
Host: Shamini Bundell
Woah, this sounds worrying. Are you saying that there’s a sudden and inexplicable increase in the number of asteroids hitting us because this is definitely the beginning of some sort of disaster film.
Host: Nick Petrić Howe
Just when you thought 2020 couldn’t get any worse. But no, that’s not actually what it is. It’s more actually that we’re getting a lot better at detecting asteroids. So, 2020 was a record-breaking year but so was several other years beforehand because, since 1998 when NASA really started looking for asteroids that are coming near Earth in earnest, the detections of asteroids have got a lot more sensitive and a lot more accurate, and now we’re detecting many thousands every year, and 2020 – despite the pandemic shutting down operations – was a year in which we saw 3,000 of these near-Earth asteroids floating around near to Earth.
Host: Shamini Bundell
So, we’re not talking about meteors which actually hit our atmosphere and burn up, we’re talking about huge chunks of rock within the sort of area of Earth. How alarmed should I be? How near are we talking? How big are we talking? Does the fact that there’s loads more asteroids than we thought mean we’re actually in way more danger of an actually harmful collision event than we thought?
Host: Nick Petrić Howe
No, I don’t think it’s particularly unexpected like how many asteroids there were and, if anything, we should be a bit relieved because the better we get at doing this and the more accurate our trajectories for where these asteroids are going to go are, the more chance we actually have of doing something. You might remember back in the 90s there were a few films about asteroid impacts and like Bruce Willis was flown to an asteroid and blew it up with a nuclear bomb and stuff like that. That is maybe now slightly more realistic because we have a better chance of actually knowing where these things are going. In terms of how close they’re getting, there were some that got really quite close but because space is vast, still safely far away. So, there was one called VT4, which was less than 400 kilometres from the planet and was almost just skimming the atmosphere but not quite, and there were a few others that were only a few thousand kilometres away, but fortunately all these things were actually quite small. It’s the things that are larger than 150-metre, these sort of objects, that we need to worry about because that could do potentially quite a lot of damage.
Host: Shamini Bundell
And I presume we find these asteroids by pointing telescopes at them and sort of looking for them. Is there any benefit to being able to map all these asteroids other than just if something is going to kill us we’ll know?
Host: Nick Petrić Howe
Well, the article I was reading in Nature, that’s mainly what is focused on, was that we would know where things are and what’s going on, but the better our sort of detection of them is gives all sorts of interesting insights for astronomers, and I’m sure they can get a lot of information from it. But I think it’s mostly we want to know if something is going to get close to us or potentially hit us. And the way it works is it’s mostly just automated telescopes scanning the sky and then when they see something moving, they compare it with a list of known objects and then if it’s something new or something unknown then they flag it to researchers who are like, ‘Oh, hang on, we need to pay attention to this, especially if its large or particularly close.’ Once they start paying attention to it, they can sort of finetune measurements and with more observations they can work out the exact trajectory, and we’re just getting better at that and we’ve got more systems in place, more telescopes doing this, and better computer programmes that can compare these objects together.
Host: Shamini Bundell
Well, I suppose that’s good news then and hopefully Bruce Willis is on standby somewhere.
Host: Nick Petrić Howe
I’m sure he’s awaiting the call. But I think that’s more or less all we’ve got time for on the Briefing chat this week. Thank you so much for talking to me, Shamini. And listeners, if you’re interested in more stories like this but instead as an email then make sure you check out the Nature Briefing. We’ll pop a link of where to sign up in the show notes.
Host: Shamini Bundell
That’s nearly all for this week, but we do have a special treat for you at the very end of the show today. We mentioned recently that various Mars missions are currently gathering data. One of those is NASA’s Perseverance rover, and Perseverance has sent back audio recordings from the surface of Mars, so we will be playing those for you at the end of the show. Have a listen out for the winds of Mars and also the sound of the rover blowing up little bits of rock with a laser. In the meantime, as always, if you want to reach out we’re on Twitter – we’re @NaturePodcast. Or if you prefer email we’re podcast@nature.com. I’m Shamini Bundell.
Host: Nick Petrić Howe
And I’m Nick Petrić Howe. Thanks for listening and here’s some sounds from Mars.