Overall, I didn't love Day 1 - some of it was interesting but the stuff wasn't really my thing. Day 2 was amazing though.
Mary McAleese
Arrived late - missed this ten-minute talk.
John O' Keefe - The Future of the Brain
I arrived a little bit into this but saw a good bit of it. Mostly it wasn't super interesting (I'm not really interested in neuroscience), but place cells were quite cool, and the fact that mice apparently know how to sense the direction they're facing innately (they learn it before they open their eyes or get spatial experience) but can't do other navigation tasks. This is the guy who discovered place cells and got the Nobel alongside someone who discovered grid cells. He also talked about how the reason you don't lag behind reality is that your brain predicts what's likely to happen in the future and prepares you for it. There was a bit about predicting Alzheimer's risk using spatial manipulation of landscapes as well.
Kay Tye - The Future of Emotion
I had spoken to Kay Tye the night before at the opening reception in the Old Library, and she was lovely and super interesting. She was really engaging on stage and had a great Powerpoint, especially when talking about the different emotional valences people ascribe to the same stimulus, such as a sprinter reacting positively, a veteran negatively, and a gamer neutrally, to a bullet sound. She talked about how some stimuli carry an innate emotional valence (like someone yelling) while others don't (like a skyscraper), unless you add something that has a valence to it (then showed a picture of the Twin Towers on 9/11).
She talked about motivated behaviours, and how an animal wants to determine whether something is important first (salience/arousal), and if no it's neutral, and if yes it's either good (approach) or bad (avoid). Apparently a person with bilateral amygdala lesions lost her fear of snakes and spiders and her ability to recognise emotions, but still panicked on suffocation (how did this happen??).
She had a funny bit about majority voting amongst neurons meaning you can have one outcome even though not everyone agrees, like in the 2016 US Election you got one president, and also about some votes counting more than others, like with the Electoral College. She then had a lot that involved jargon-y names for parts of the brain and it lost me quite a bit, though the bit about being primed for escape but not for reward was interesting.
 |
| A crude diagram - sadly I did not take down the legend so it's a bit of a mystery now. |
So yeah, lovely person but the topic wasn't really my thing I think, and a lot of it went over my head.
Danielle Bassett - The Future of Complex Systems
This one was definitely a new one to me, but I will give it one thing - for all the things I've heard promoting STEAM (science, tech, engineering, arts, and maths), this was the first time I'd heard of the arts actually being useful in answering a scientific question directly, and it was to understand the brain and the brains of people throughout history by analysing old works of literature and art computationally, looking for clues about brains by looking at the work they produced.
Ada Yonath - The Future of Structural Biology
Yonath won the Nobel for discovering the structure of the ribosome, and that's what most of her talk was about. It was really interesting and she was very charismatic, talking about new targets for antibiotics, how they made their discoveries, her grandchildren, and the origin of life (she says it came from the protoribosome, an extremely conserved central segment of the ribosome containing autocatalytic RNA). A very good talk.
Beth Shapiro - The Future of Extinction
Shapiro was probably the most engaging speaker of the conference, though there was some competition from people like Kay Tye, Nick Lane and Emma Teeling. She talked about the three strategies we can take to extinction - normal conservation efforts, genetic rescue (stuff like figuring out inbreeding and who you can outcross a population with, or gene editing like with American chestnut trees being edited to make an enzyme that breaks down the oxalic acid a fungus kills them with), and de-extinction, i.e. things like bringing back woolly mammoths. (She does not agree with de-extinction, one reason being the lack of places for the animals to live). At the start she read out a funny and very TED-talk friendly hate email she'd received shouting at her because her 'How to Clone a Mammoth' book did not in fact teach people how to clone mammoths and now this person's daughter was going to be very disappointed.
The topic isn't one of my core interests but she is definitely very engaging and it was a good and accessible talk.
Svante Paabo - The Future of Ancient DNA (Research)
I just couldn't get into this one unfortunately - I'm not very interested in ancient humans, and I was also very tired and hungry by this point. Such is life.
Leroy Hood - The Future of Medicine
I'd been interested in this guy because he invented so many of the tools molecular genetics is built on, like the automated DNA and protein sequencers and synthesizers, but he talked about his P4 medicine (predictive, preventive, personalized, and participatory) and, while I agree with that and making wellness a focus of medicine as well as sickness, his presentation seemed very Silicon Valley, like a slide deck for a pitch, rather than a scientific talk.
Feng Zhang - The Future of Gene Editing
I'd been super interested in this because this guy co-invented CRISPR and I was reading the book by Jennifer Doudna about CRISPR at the time of the conference, so I was looking forward to hearing about the implications of gene editing for the future and the potential uses it could find - but he just talked about his work engineering the system to edit RNA molecules. It's certainly interesting, but I guess not what I was hoping for given the title of the talk. I did like his approach though, of using bioinformatics to find candidates and then doing the wetlab work.
Karl Deisseroth - The Future of Brain Editing
This was the talk I was rapporteuring on (that's totally how words work). Deisseroth invented optogenetics, which allows you to turn neurons on and off using light, and CLARITY, which makes a transparent copy of brains, and is a researcher and psychiatrist. He talked about how optogenetics works a little and about depression (I guess as a potential application?). They can make it respond quickly or redshift the colour the neurons respond to.
This quote: 'It may not be very useful to paralyze a worm, but we can definitely do that, so if anyone needs a paralyzed worm...' They can also apparently mind-control worms, making them not act thirsty even when they're dehydrated, or lick like they're thirsty even after they're sated, and investigated how well they can control them when they're satiated vs not. I feel like the ethics of this are very, very bad. But yeah, they can mind-control living animals at the resolution of each particular neuron.
Murray Shanahan - The Future of Artificial Intelligence
I was very tired by this point unfortunately but basically he talked about neural networks. I remember one of his final points - the dangers of AI aren't likely to be Terminator-style destroying humanity to protect itself, because that's anthropomorphic (though if it became conscious it might develop a sense of self-preservation, or want to preserve itself to continue its task?), they're more likely to be like in the Sorceror's Apprentice with the maliciously compliant brush (let's hope I'm recalling the plot correctly - I read it as a small child).
Saul Kato - The Future of Computational Biology
On the definition: 'had to look up what computational biology is, consulted the oracle of human knowledge, wikipedia' but he thinks it's 'biology when you can't just look at it'. All the data we have is 'a huge exploding way of keeping biologists employed today'.
He works with the nematode C. elegans, and talked about the map of its 302 neurons that took decades to create but apparently hasn't been that useful: it shows that 'there are a lot of genes, and they have relationships, and I'm not sure what else you're gonna take away from it'. He says that network structure is insufficient to understand biosystems, because biological systems are dynamical and have a function. 'The bad word is teleology...If you wanna talk about function now talk about teleonomy, no one will think you're a creationist'. He talks about how we need to look at input-outputs and time series to understand biological systems, and says 'the future of computational biology is more data -
not just because of the need to stay employed'. So yeah he's all about time-series data during development or during movement through the world.
He's a funny guy, and also seemed enthusiastic, so I quite liked him. Even if he was a physicist originally.
Daniel Dennett - The Future of Life
DD is a big-name philosopher and got an extremely swooning introduction, which was a bit much I feel.
His talk was very rambling - it had some bits that resonated with me, but also quite a lot that didn't, and it very much was not about the future of life. That was a recurring theme - speakers not sticking to the titles they were given.
What I liked: his comparison of a termite mound to a Gaudi castle it looked extremely similar to, and the discussion of humans as intelligent designers, and how we get from uncomprehending but competend bottom up design to top down intelligent designers. I also liked his description of evolution as blurting something out and then making it better, because you can't edit what isn't there, which is a motto I originally learned from writing. The Darwinian-Skinnerian-Popperian creatures thing was cool too, though Gregorian seems like an unnatural addition. Basically it's like Darwinian creatures try things and may die, Skinnerian creatures do things and get a reward or don't and learn, and Popperian creatures have a selective inner environment so we try things out in our head and have a higher chance of making a good first move (may not get a chance to try a second time).
It was fairly inconsistent though, and the intro definitely seemed overblown.
_______
In general, the day was a bit disappointing, though there were some good points. Also, I walked back to college with Aoife, Claire (a journalist I know), Dara O' Briain and a couple of others, which was an interesting experience. The second day was a lot better.
No comments:
Post a Comment