From the Original Series to Star Trek: Picard, Siskoid and Scott X trace the history of androids and the story-telling strategies used to make them mirrors held up to humanity. What are androids, what do they mean, and what's the next step in their evolution? Plus, Star Trek News and your feedback from the previous episode.
Listen to Episode 41 below!
Or subscribe to Gimme That Star Trek on iTunes!
This podcast is a proud member of the FIRE AND WATER PODCAST NETWORK!
- Visit our WEBSITE: http://fireandwaterpodcast.com/
- Follow us on TWITTER: https://twitter.com/FWPodcasts
- Like our FACEBOOK page: https://www.facebook.com/FWPodcastNetwork
- Use our HASHTAG online: #FWPodcasts
- Support us on Patreon: https://www.patreon.com/fwpodcasts
Subscribe via iTunes as part of the FIRE AND WATER PODCAST NETWORK.
Credits:
"Star Trek Theme" by Alexander Courage, with the Irredeemable Shagg on vocals. End theme: "Deep Space Nine Theme" by Dennis McCarthy.
Bonus clips from: Star Trek: The Next Generation's "Encounter at Farpoint" starring Brent Spiner, "The Offspring" starring Brent Spiner and Hallie Todd; Star Trek's "Requiem for Methuselah" starring William Shatner and Leonard Nimoy, "I, Mudd" starring Leonard Nimoy and DeForest Kelley; and Star Trek: Picard's "Et in Arcadia Ego: Part 2" starring Jonathan Frakes and Brent Spiner.
And thanks for leaving a comment!
Podcast: Play in new window | Download
Thought provoking as always. Lots of interesting questions raised. I just hope Picard Season 2 rises to the challenge and addresses some of these, as opposed to hand-waving it away. Nice to hear they don’t plan to, at the very least.
Always a pleasure to hear more from Scott X!
Chris
Great episode, gentlemen. I really enjoyed the discussion.
As a biologist, I come at the topic of sentience from the perspective of animal cognition rather than AI. I particularly liked the definition that Scott presented, and the idea that we only understand sentience (as much as it can be understood) as it has evolved in hominids. Other species may have evolved different forms of sentience, or sentience may exist along a continuum in different species, rather than being a binary presence/absence kind of thing. In the end, that leads me to believe that we should be extremely careful when deciding whether something is sentient and deserving of special rights, or whether it is merely a biological or artificial machine that does not deserve such rights.
On a slight tangent, I’m reminded that we humans tend to think of sentience as the next best thing to sliced bread. In the end, it is just one of many different adaptations that living organisms have developed to be evolutionarily successful. Only time (and the forces of natural selection) will tell if sentience is really as great as we think.
Always a pleasure to hear more from Scott X!
What Chris said.
Hey Siskoid! Great episode. Though I wouldn’t call myself an expert on artificial intelligence, as an IT product director and computer engineer, I do have some first hand familiarity with the subject, so I thought I would give some background and context. The starting point of all AI is algorithms and data. The goal of AI is to create a set of self-correcting business rules, combined with a predictive model to consume relevant data and achieve a desired goal.
I’ll give an example: Let’s say that I want to sale t-shirts. I want to give the customer the color t-shirt that they would most likely like. I sale two colors: blue and pink. A simple (and admittedly sexist) algorithm would be if customer is a boy recommend blue, if girl, recommend pink. This would result in a computer program, no matter what, recommending blue for boy and pink for girl. It is incapable of making a recommendation outside of the binary choices or for anyone who identifies as non-binary.
The first step in AI would be to provide statistical data and to record the efficacy of the computer’s decision (I now add to my website a “Do you like this choice” button). Now the algorithm looks at how often a person of a specified gender chose which color and recommends based off of how often it has been selected by a like person in the past. Bias can still creep in if the data itself has been impacted by bias. In the resume selection example Siskoid gave, the algorithms could have been biased, but so could the data. If it took into account how often people with certain inputs on the resume were hired, and the company always had an in-person interview which determined who got hired, any biases of the person making the hiring decisions are inherited unknowingly. This is where most AI sits today. A lot of populist press presents AI as the computer replacing the human, but their still a human deciding the goal, what parameters to consider, and monitoring the outcomes in order to tweak the algorithms. The AI becomes more complex the more data I feed into it. Age, income, zip code, etc. This is why “big data” has become such a hot commodity. Where the power is in AI as it is most commonly used is as the data becomes more diverse, the decisions become more precise and more individualized. I work in Kidney Care so the difference between an algorithm which always recommends a dose for anemia therapy based on a patient’s most recent Hgb results vs. once which is recommending a dose based on how the individual patient has responded to anemia therapy in the past is really powerful.
The AI becomes more complex the more data I feed into it. Age, income, zip code, etc. This is why “big data” has become such a hot commodity. Where the bleeding edge is of AI is AI learning to make new connections with the data. Does color preference wane based on seasonal affect? This still relies on what data is chosen to provide to the artificial intelligence, and the goal provided to it. You see this in War Games. “Joshua” takes what he learns from the behavior of Tic-Tac-Toe and applies it to other games (i.e. War) and comes to the conclusion those things which were defined previously as winnable, are in fact unwinnable. An artificial intelligence cannot learn history if it is only provided biological data, nor would it try to. We do not have (at least that I am aware of) AI with self-determinant goals or programmed “curiosity”.
What a show on Androids and no attempt to answer the question Do Androids Dream of Electric Sheep;)
One aspect with Androids that wasn’t talked about that could be interesting is playing with the ship of Theseus idea.