Home Creators Posts Import Register Favorites Logout

Content

Hello lovely Patrons!

The new episode premieres TONIGHT at 20:00 British time - I'll be in the live chat from 19:30 so you can come and watch it with me! Can't wait to see your reactions!

Abi XXx

Files

What Ethical AI Really Means

NEBULA: go.nebula.tv/philosophytube Subscribe! http://tinyurl.com/pr99a46 Twitter: @PhilosophyTube Instagram, TikTok, Tumblr, BlueSky: @theabigailthorn Facebook: https://www.facebook.com/PhilosophyTube/ MERCH: https://store.nebula.tv/collections/philosophy-tube Email: philosophytubebusiness@gmail.com MUSIC: Readers! Do You Read by Chris Zabriskie is licensed under a Creative Commons Attribution 4.0 license. https://creativecommons.org/licenses/by/4.0/ Source: http://chriszabriskie.com/reappear/ Artist: http://chriszabriskie.com/ Original Music by Nina Richards: https://www.ninarichards.co.uk/ BIBLIOGRAPHY: https://docs.google.com/document/d/e/2PACX-1vTw_9TLluve5BJonaXGaon2QrCJYcORTZlN4EOY5ebhv6jleTGQ5pDOPZTZrlJcfqF82qntyITnkyn3/pub 00:00 - 11:38 Building Ethical Computers 11:38 - 20:42 AI in An Unjust World 20:42 - 35:53 Using Data Responsibly 35:53 - 38:47 The Ballad of Kelly Slaughter 38:47 - 49:34 AI is a Physical Thing 49:34 - 56:38 Conclusions #AI #Science #Philosophy

Comments

K

I'll be cancelling my Patreon subscription and didn't want to do so without explanation. You've built a sizeable platform from producing content on social justice, ethics, etc and when it really matters you're silent about the genocide of Palestinians, despite your large platform. Of course if you do decide to get on the right side of history I'll be pleasantly relieved but as of now I feel you've revealed your true colours and I don't feel comfortable contributing to your Patreon, regardless of how relatively small my contribution was. This is a real shame for me as you're the first content creator I made a Patreon account to contribute to.

Jay Kamins

K, I share your concerns about the situation between Israel and the Palestinians and the innocent people suffering on both sides of this senseless conflict. That said, it seems a bit unrealistic to expect Abigail to be giving real-time commentary on every pressing topic going on in the world. Wishing you peace.

Reggi

Imagine asking someone to wrap their head around a 100+ year-long problem and make a statement about it, isolating part of their fanbase for $2. Now imagine requesting that when your local and national politicians haven't done that work, the place where you bought breakfast hasn't done that work, and whatever megacorp owns, your favorite drink hasn't done that work. Wishing you solace farming and harvesting rainwater. -_-

K

Not asking her to do anything for my money. Just thought it would make sense for me to explain why I'm withdrawing my money. It doesn't take much for a celebrity who brands themselves as caring about social justice to realise genocide is wrong. That's not my jam and I'll leave her to that part of the fanbase she's supposedly worried about "isolating". I'm not in support of those megacorps any more than I need to. I don't need to support Abigail if she's not who I thought she was when I signed up to her Patreon, that's all.

Amyphist

Exhibit A for 'you cannot please everyone'...

Joshua Rodman

You know that Philosophy Tube has months long leads times for the generated content, right?

Joshua Rodman

Hey, here's my comment irrelevant to the video. In light of youtube's war on adblock, it would be great to have a place where patreons (patrons?) can view your work elsewhere. I don't know if there's a good option for you there but expressing the interest. (For me, you're already good as I watch on nebula.)

Declan

Currently running a small recruitment company implementing some AI solutions (mostly as administrative aides as opposed to being part of the selection process) but this video makes me feel yikesey.

Michael Masiello

That’s it. I’ve seen enough from this channel. I had no choice but to become a patron. Abigail, you (and your hard-working team) are amazing, and I thought this take on AI was as nuanced and trenchant as it was artful and clever (i.e., very). Really proud to support what you do.

Kt

This episode is one of a long line of impressive episodes. Your work is great from the production design and quality, which catches the eye, to the concise but wide ranging philosophical basis for each video. It all flows together so well. It's visually gorgeous and subsequently it's a little horny for audiences in the best way possible. Plus the sex jokes are fucking A++++. Can't stress this enough honestly. But flowing throughout and underneath it all is a fascinating concept that ties everything together in a unique package that no one else can provide. This is why I watch your show.

AmiAndTea

This was a great episode, weird to hear Abi cite someone I know through my day job though. Fabian’s done some really cool work on deep machine learning (the types of information processing people often mistake for AI) but I didn’t know he had published any writing on the topic - and now I’ve got a new peer reviewed journal to browse for my work reading! 😄

Noblese Oblahaj

"Come with me if you want to live." "... in general, or in Britain?"

hakugei_

I'm french and one of the public services here that gives benefits is using algorithms managed by a private company (so not open-source) that have been creating kafkaesque situations. You can get your benefits for months and one day, out of the blue, a letter saying there was a mistake and you have to give part of (or all!) the money back. With no explanation, no way to check the calculations whatsoever. No way to know if maybe you did a mistake when you filed, or if the thing is just malfunctioning and asking for money back when it shouldn't. You just have to take their word for it and give it back. Newspapers talked a little bit about it for a while then nothing. And to my knowledge, the problem didn't stop. So much for "the country of human rights"... I was entitled to some benefits for housing during several months when I was unemployed. During this period, my income didn't change, my rent didn't change, but I was getting a different amount each month. Never knew why... And I had to live with the fear that I would have to give it all back at some point. It felt like playing at the casino against my will. Fun!.. Honestly, watching your videos just helps me feel less alone: I work in tech now and most of my colleagues love Dall-E and Chat GPT and use them all the time. Many of them even subscribed and pay 20$ a month for Chat GPT. It's so depressing when I have to put up with conversations around this subject... When I try to bring up ethics, I just feel like the annoying killjoy in the room that just can't appreciate something fun. I tried once to get some of them to watch the documentary "Coded bias" but none of them did eventually. I've stopped trying... I also wanted to thank you for the ever growing quality of your content, I've been following your channel for a while now and it's very impressive. You all do an amazing job, and this kind of content is much needed and appreciated.

Tarot

Hi there! I'm not sure if you cover this in the video, I'm still watching it, but I'm in publishing and this is something that came up very recently. I know you cover screenwriters and playwrights, but there's a lot of trouble with the AI groups also taking information from copyright-protected published novels (grabbed through pirated content, of course). Here's the link to the Author's Guild and their lawsuit against OpenAI: https://authorsguild.org/news/ag-and-authors-file-class-action-suit-against-openai/ (I'm also interested in the legal side since I am applying to law school to study intellectual property). If you didn't know about this, just more info for your points!

Michael Masiello

I mean, does K here think this gets done overnight? All the emphasis she placed on how much time and effort goes into these things—totally ignored. And if she did tackle this subject, one suspects she’d do her homework first, though the intimation that someone who cares so deeply about human suffering would tacitly or overtly condone genocide is really just…well, astonishing.

K

I'm not gonna dignify most of these replies with a response - Regarding the lead time to produce content - where did I say I expected her to make a video on this? She's one of the few public commentators I follow on social media who wasn't using their social media platforms to call out unprecedented (since Iraq) levels of media dishonesty, advocating and sharing about demonstration in solidarity with the Palestinians or at the very least sharing fundraisers for MAP. A retweet or statement isn't hard - and silence is a historic week like this was a statement of it's own. And yes you're right the implicit condoning would be astonishing - hence my decision. Silence is complicity. For many of us, the Palestinian people have been all we can think and talk about this week - so forgive me for being disappointed in the silence from a channel I followed because I thought I had shared values. Even if Abi needed to still "do homework" on Palestine (which is ridiculous - it's plain settler colonial genocide anyone can see that) - that's still yeah not someone I'd want to subscribe to if they're not sure where they stand on the eradication of a people. Not a judgement on her as much as it is on why is she's the one I'm Patreon subscribed to when there's that misalignment. The past week has seen some of the worst genocidal rhetoric and actions in my lifetime - "doing her homework" isn't acceptable at this point to me, nor is that what she's needed. She has actually now made it clear on twitter that she's aware something is happening in Palestine and has some kind of stance - but in an almost cowardly (or just in a genuinely "both sides"-ing a genocide liberal-minded) and wishy washy way that I'm personally not happy with. That's fine, she has the right to her opinion (and whatever criticism I have of the kind of angle she's taken, it's a league better than most of the pro-genocide shit out there) and doesn't have to speak out on everything. I'm not saying anything abt whether others should support her or that I have anything against her. Just her silence during this week was noted. I'm not sure why everyone's so annoyed abt my comment as I've said I'm 'cancelling' her - her content is still great and I'd still recommend it to others. Just I'll no longer be a subscriber (once my yearly prepaid membership expires). Hardly a career-ender for her lol - and not a judgement on her fans.

Stormsdream

Wonderful explanation for "hey, maybe don't post fan fiction online". A lot of effort, sacrifice, blood, sweat, and tears does go into making stuff, and published/posted fan fiction in all of its various forms kind of . . . doesn't care

Diana Godin

Do the episode footnotes/citations get posted anywhere?

skytwosea

You are the best. Phenomenal episode, incredibly informative.

Michelle D'israeli

I'm a cyber security professional, and this video really should be required watching for anyone in my industry. Of special note is the conclusion, which mirrors something I keep pointing out - all the risks around AI are the same ones we've always had, they're just magnified. From faking images to spreading disinformation to being able to generate meaningless dropshipping advertisement blog content, all this is an issue right now. Guessing Abigail originally had a section on this, but it was cut for time and/or flow reasons: Companies de-risk Data labellers by comparing their entries to those given by other data labellers. This alone means that they have to not label correctly, but label according to what they believe others would pick. Add to that the need to complete many thousands of labelling exercises a day in order to earn anything, this ultimately means that data labelling is no longer actually as useful as it should be. Data labellers no longer label according to quote-unquote reality, but according instead to "what would other data labellers pick within less than 5 seconds". I'm sure there's a whole bunch of interesting philosophical concepts in that just in and of itself :P

Ish The Dark Lotus

The thing that is said in the end is the same problem I face when speaking about dangers of AI to regular (non-tech/non-activist) folks. They only care about Skynet and dangers of 1990s hollywood-movie style AIs. As soon as I talk about real-world problems of AI involving surveillance, exploitation and prejudice, their eyes gloss over and they lose interest. One strategy I'm thinking of trying is doing what Abigail did in this video, aka, start off by humoring Skynet/AGI to get people's attention and then gradually take the discussion towards real-world AI problems. Like - "Yes, Timothy, a future AI overlord ruling over humanity is a scary idea indeed. In fact, depending on the training set errors, a good AI overlord can easily turn into an evil AI overlord." Now, when Timothy is invested, I can do - "As an example, did you know, that even currently, when training models of face recognition include only white faces, the computer cannot distinguish people of other races well, and this can lead to an innocent person being framed as a criminal. Currently, many law enforcement agencies already do this, and if we don't stop them now, imagine the dystopia when Skynet takes over and tries to enforce law and order in human society, how terrible that would be."

Ish The Dark Lotus

Ouch, so even at the labeling stage, there are chances of conformity bias. This is really insightful.

aura

Just subscribed. I am a researcher working on AI and I think this video and/or a LOT of the references in here should be part of any serious curriculum on AI. You are giving us very good material to raise awareness here.

Kerry Luckett

This is another masterpiece! I teach in the communications department of a two-year college in the U.S., English and ESL, and my colleagues and administrators are overcome about students using AI to "cheat". But they do not really understand what "artificial intelligence" means. Once upon a lifetime (90s to early 2000s), I worked in tech and saw the evolution of a lot of these "large-scale computing" products. Many of my co-workers then were impressed. I was terrified because I could somewhat see the direction the products were going. This episode deftly points out all the very human problems in computing in a digestible way. Thank you, Abigail. I am going to share this with select colleagues because we need to educate ourselves about this real and present threat.

B

My stomach definitely turned when a recent assignment asked us to program style transfer between a landscape photograph and various popular paintings (e.g. Starry Night).

Kristen

It's funny, because my "diverse" team is so dysfunctional that GitHub Co-pilot (an AI for coding) is more helpful with day-to-day coding issues than the people, but it can't solve the bigger issues of non-communication because everything is different services talking to each other, so while the code is important it can't do anything without knowing the right inputs and expected outputs. Yet, everyone carries on like nothing is wrong and we're all very smart and hard-working and happy to be in a "skilled" profession with a "career." There's a facade in tech that is so deep, and so obvious when you look at the propogation of the internet (no one is or was working 8+ hours a day with cat pictures or open source projects a click away) that I'm sad I fell for it and went to study computer engineering instead of continuing with philosophy. But, y'know, girl's gotta eat.