Can A.I. Save The Middle Class? | EP 77

Подписаться 21 тыс.

Просмотров 19 тыс.

50% 1

This week we look at how A.I. is affecting jobs. As companies start announcing A.I.-related job cuts and experimenting with customer service bots, economists are placing bets on whether A.I. will lead to major gains for companies and workers. Some are even predicting it will help rebuild the middle class.
Then, the multidisciplinary artist and filmmaker Paul Trillo joins to talk to us about his experience as part of a select group of testers granted early access to Sora, OpenAI’s video generation tool. And finally, Kevin explains what happened when a Microsoft developer stumbled on a huge cybersecurity breach.
Chapters
00:00:00 - St. Stupids Day
00:00:56 - this week on hardfork
00:01:38 - A.I. and jobs
00:22:30 - Using Sora with Paul Trillo
00:47:01 - Secret Software Attack
Today’s Guests:
Paul Trillo, multidisciplinary artist, writer and director
paultrillo.com/
Additional Reading:
How One Tech Skeptic Decided A.I. Might Benefit the Middle Class
www.nytimes.com/2024/04/01/bu...
Will A.I. Boost Productivity? Companies Sure Hope So.
www.nytimes.com/2024/04/01/bu...
Paul Trillo’s Sora Video, “the Golden Record”
vimeo.com/927400626
Did One Guy Just Stop a Huge Cyberattack?
www.nytimes.com/2024/04/03/te...
Hard Fork is a weekly look into the future that's already here. Hosts Kevin Roose and Casey Newton explore stories from the bleeding edge of tech.
Casey’s publication, Platformer: www.platformer.news/
FOLLOW US ON SOCIAL
TikTok:www.tiktok.com/@hardfork?is_f...
Insta:
Kevin Roose- kevinroose...
Casey Newton - crumbler?h...
Twitter (X):
Kevin - x.com/kevinroose?s=20
Casey - x.com/CaseyNewton?s=20
Threads:
Kevin - www.threads.net/@kevinroose
Casey - www.threads.net/@crumbler
Subscribe to the audio only podcast:
Apple - podcasts.apple.com/us/podcast...
Spotify - open.spotify.com/show/44fllCS...
Amazon - music.amazon.com/podcasts/7c7...
Google - podcasts.google.com/feed/aHR0...
The New York Times - www.nytimes.com/column/hard-fork
Credits
“Hard Fork” is hosted by Kevin Roose and Casey Newton. Produced by Whitney Jones and Rachel Cohn. Edited by Jen Poyant. Engineering by Alyssa Moxley and original music by Dan Powell, Elisheba Ittoop and Marion Lozano. Our audience is Nell Gallogly. Video production by Ryan Manning and Dylan Bergeson. Motion graphics by Phil Robibero Thumbnails by Julia Moburg, Elizabeth Bristow, and Harshal Duddalwar
Special thanks to Paula Szuchman, Pui-Wing Tam, Nell Gallogly, Kate LoPresti and Jeffrey Miranda.

Наука

Опубликовано:

14 июн 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 78

@hardfork 2 месяца назад

Are you using generative AI at work? Tell us about it in the comments.

@x3haloed 2 месяца назад

I use it very often in software development. Think about it this way… software development has a good measure of need to just apply known good practices to common problems, BUT that’s not really the essence of the job. The essence of the job is to apply creative problem solving to novel problems over and over again. The real meat of doing development work is that your problem is often similar to other solved problems, but almost never identical. This is where generative AI comes in. It’s like a thinking search engine for all known engineering solutions. You can ask it about solutions to problems like yours and compare and contrast them. This can kind of be done already by searching StackOverflow, but what stack overflow cannot do is find results that are “kind of like this one situation, but slightly different, and also I don’t know the exact name for what I’m thinking of.” StsckOverflow also cannot take a given solution and adapt it to your own situation. In some situations, this has reduced what could take several days of research for me down to just several minutes.

@scoty_does 2 месяца назад

Automation has been taking jobs for 20 years. Maybe it just continues a the same scale. People that make AI tools are over selling it ?

@RareBirdGames 2 месяца назад

We are using it to help conceptualize things faster so the people on our team making content know exactly what the target is, and gets them on the same page faster rather than iterating for weeks. We see AI being used to run all the underlying bureaucracy that keeps our society turning, freeing us up from having to "hustle" to survive and letting people have the time and space to achieve whatever they want in life. We hear a lot of people worried about the AI movies and music, but you can opt out of the internet and a lot of people are excited to do so, they just are forced to have social media to survive. This current system isn't working for most today, and we need to fully embrace it and skip the prohibition phase that we are always excited to jump to. We just have to accept and adapt, so we can start benefiting from it sooner.

@evdm7482 2 месяца назад

I mean, let’s just stop technology, we did it, done, let’s all go home and call it a win and never do anything again. Or… the future. Good luck.

@marcoshoracio 2 месяца назад

I'm a film editor and producer and I've been using Midjourney, Elevenlabs and Pika to create "sizzles", which are pitch/concept videos for documentary projects we're trying to take off the ground.

@stanakar 2 месяца назад

Don't want to mess up your Spotify numbers, but you told me to come here for the visuals today!

@laurenwhicker6612 2 месяца назад

I’m normally an Apple Podcast listener, but I came here to say the music was FIRE this week and also I didn’t realize you record together in studio?? Am I a RU-vid convert now?

@theoaklandpanda9129 2 месяца назад

I like turtles

@princesssshortie 2 месяца назад

Same! I figured they were over zoom or something. Also, they don't look like what I pictured. Haha

@cassiophone 2 месяца назад

As a usual audio only listener, I am disappointed to learn you have faces and are not in fact AI.

@smaranda.p-p 2 месяца назад

😂

@jupiter75100 2 месяца назад

Long time listener, first time watcher🎉

@williamminer2826 2 месяца назад

I've only ever listened to the podcast before now and Kevin talks with his hands more than I pictured

@RyanRuark 2 месяца назад

The idea that AI will do anything other than continue to extract wealth from the bottom and funnel it to the top is the height of techno narcissism.

@Earthwirm 2 месяца назад

How does that work again? If top 1% pay 43% of the taxes in the US.

@dwallace67 2 месяца назад

yup. they are stuck in a delusional fairy land

@montymonto6430 2 месяца назад

@@EarthwirmHow does it work, what percentage of wealth does the top 1% own? Don't just parrot FOX news. Also how much of the decisions are made by the top 1%, about 99%? How does that work?

@mikezooper 2 месяца назад

@@EarthwirmThe top 0.1% don’t pay tax.

@x3haloed 2 месяца назад

That’s some thinking with finality to it. Good thing we’ve arrived at the end of machine learning, so we know exactly how the story ends.

@rachel_rexxx 2 месяца назад

Holy cow, Sora as it is might be perfect for making lookbooks for ads

@Profdesoussol 2 месяца назад

why oh why did I watch the podcast on youtube. I saw your faces. It's not that I'm disappointed, it's just that I had my own version and now it's broken forever

@Brian-oz8io 2 месяца назад

I think once companies figure out exactly what kinds of productivity standards are going to come from using ai, it’ll be harder for employees to use it to do less work. Instead of “Bob, we need 3 TPI reports today” and Bob just uses ChatGPT to do them in 20 minutes, it’ll be “Yeahhhhh, Bob we know you’re using ai and the average ai assisted worker can do 75 TPI reports a day, so you’d better get to work because we want all those on our desk by the end of the day”

@lifebasesolutions3405 2 месяца назад

What would ya say... ya do here?

@desertislanddivs 2 месяца назад

The shock jocks of tech. Totally ^^ You guys are the best Hat GPTs out there.

@lincylinc 2 месяца назад

I love the podcast. Haven’t been able to wait until the RU-vid video is online. Now I want my cake and eat it too.

@smaranda.p-p 2 месяца назад

Kevin and Casey, You're great! Always making me laugh 😄 Thank you for another entertaining and interesting episode!

@thehappyhound770 2 месяца назад

Kudos to Casey for holding the back door joke until it really counted!!

@AP-lb3ot 2 месяца назад

Guys...PLEASE add an Epilepsy warning to the start of this podcast. I just about spiraled.

@mygirldarby 2 месяца назад

Really? They should just change it if it causes anyone a problem. The Epilepsy Foundation says that is extremely rare. Only 3% of people with epilepsy have seizures that are triggered by flashing lights, and they are usually children/adolescents. 97% of those with epilepsy don't have that issue. Still if their intro is causing anyone problems, they should change it.

@stephaniew6020 2 месяца назад

Hi. You guys got at a question about AI in a few of the segments that I've been asking myself a lot but haven't known where or how to ask. And that is, what do we lose in terms of human development with a reliance on AI? The answer went a couple of different ways during the course of the podcast. In a couple of instances, people recognized that they wouldn't want to completely rely on AI because they understand that going through the motions and doing the actual work is important to our own learning and satisfaction (one was the real estate guy, the other was the filmmaker). I don't hear many people talking about AI in that way so will look forward to future episodes where HF gets into that question more deeply.

@stephaniew6020 2 месяца назад

Also, the question about using AI at work, I work in public health at the state level and recognize that access to information is a social determinant of health. Most of the info we put out is framed for a specific audience that we call "the general public." Beyond language translation, which I think AI can probably assist with, how can AI help to capture/address specific cultural framings of public health issues? We haven't gotten there, yet, but I'd be curious about people who have or if anyone is thinking about this. My suspicion is that because AI is biased towards the dominant cultural paradigm, it can't help that much.

@kangarool 2 месяца назад

no offence but would have preferred the film guy's entire segment to have just shown me the whole Baby Petri Pandas movie, they're chemically ADORABLE!

@tessapulaski3592 2 месяца назад

Long-time listener, but first time viewer! I enjoyed watching with the visuals, but also threw me off a bit.

@lnrefn 2 месяца назад

I enjoy your podcast. You mentioned NPs as 'mid tier" care provider but when you see one, do they give "mid" care, or do you feel half or mid better after receiving care ?

@BruceWayne15325 2 месяца назад

I use AI every day at work. It's a nice performance booster. It's a brainstorming beast. I'm able to throw together a plan in 1/20th of the time, but it's not going to take many jobs for a while. It's too unreliable. It would be like hiring someone with dementia and Alzheimer's rolled into one. If that person could do your job, then you best worry. Otherwise, you're going to be safe for a while. Once they make it reliable, which I do believe will happen, then you'll see mid-high double digit unemployment. That's not something that worries me though. We're headed towards a free labor economy. We're going to have to redefine our entire concept of economy at some point.

@jacqdanieles 2 месяца назад

In your estimate, when will it become reliable?

@mygirldarby 2 месяца назад

@jacqdanieles I'm an AI trainer and AI is more like a very precocious 3-year-old than a person with dementia. If it grows up quickly, and I think it will, we will no longer be able to guide it within 6 years.

@BruceWayne15325 2 месяца назад

@@jacqdanieles that's the million dollar question. As long as they keep relying on deep learning I think they will continue to have issues with reliability, though it will improve significantly. If we want it to be completely reliable then they need to switch to cognitive, self-learning, self-adapting models. When a machine learns like a human, then I suspect we will be able to see reliable AI. They can still use deep learning to train a knowledge repository that the cognitive AI accesses like a human would use ChatGPT, but the cognitive model would use reasoning, logic, and research to determine truth.

@dodgestregis4990 2 месяца назад

@53:00 - feels like those dudes in the cold war that just because of them we aren't all dead now

@conzure 2 месяца назад

New music? It rocks

@human_shaped 2 месяца назад

The next generation of kids will be called "Gen Gen". I've heard it in a few places now. I like it.

@michael7018 2 месяца назад

I use AI at work as a personal assistant as part of my role - Cybersecurity Risk Management, I use AI to line up risk frameworks like NIST to document compliance and remediation requirements to address risks and more. The this is, you can’t replace an expert in that field because you need to put the AI results in context. So I agree,AI will have a positive impact overall to save time with the development of reports, codes, presentation, etc. I don’t believe AI will kill a great number of jobs. Maybe a few but overall it’s going to make enterprises more efficient to focus on the most important things.

@drmarioschannel 2 месяца назад

similar debates were had when the internet emerged. we already know the ending.

@isaacliu896 2 месяца назад

David Autor will become Paul Krugman fax machine famous

@theelderskatesman4417 2 месяца назад

No. It will make obscenely rich people even more obscenely rich as their AI systems absorb and remix the output of actually creative people. Anyone who expects anything else is deluded.

@alecfotsch3533 2 месяца назад

Is the golden record actually what you hoped to make? It's definitely dynamic. Also nauseating.

@jankoaltenschmidt7363 2 месяца назад

Whaaaaaaaa - how is it possible that Kevin's takeaway from the XZ distaster is that all software needs to be open source???!!! Open sourcing is the very reason how a malicious actor could sneak harmful code into a product that gets maintained by unpaid and overtired individuals! Open source AI means that all of the investments in safeguards by the proprietary model providers are consider null and naught by the open sourcers as any actor can strip them away on the source code 😶‍🌫️

@kbermond 2 месяца назад

When does the Linux Backdoor Bandit #Netflix series premiere?

@dodgestregis4990 2 месяца назад

When you have the a three shot of you two and the guest you need to put the guest in the middle

@dodgestregis4990 2 месяца назад

The split screen

@late_privktorian_era 2 месяца назад

Why

@dodgestregis4990 2 месяца назад

@@late_privktorian_era because the guest is the center of attention

@dodgestregis4990 2 месяца назад

And it's better for the edits and cuts to guest only statements

@emmafreeman881 2 месяца назад

I don't follow the argument for open source being more secure. It seems that the back door to linux was possible thanks to the manipulation of a tired volunteer and discovered by a paid employee. I'm not saying I love massive tech companies, just not following the argument at the end of the episode. Surely paying people for their time has some important security implications.

@brandonreed09 2 месяца назад

What if Gia Tan was actually the NSA?

@tangerine8912 2 месяца назад

When you write a piece and AI points out the subtext and you correct it to make it sound a certain way, aren't you taking away the human emotion out of it? Why not share your original thought even if they are not "perfect" as per AI standards, at least they are original and human. What is the reason for updating it?

@hallohallo3412 2 месяца назад

Feedback: Podcasts are cool because you can listen them while doing other things. Makes it a little harder if you have to watch too.

@weredragon1447 2 месяца назад

Awww...no more "TPS" reports? Bummer. 😂😂

@joshharper7466 2 месяца назад

You guys look much cooler than you sound

@nathanielbrewster8457 2 месяца назад

Weird voiceover edit around 3:30

@TeddyLeppard 2 месяца назад

What middle class?

@TheManinBlack9054 2 месяца назад

to hell with middle class, why no one cares about the working class?

@ili626 2 месяца назад

20:25 This is the same guy, who not long ago, thought GPT-4 was in love with him and might be conscious. He went from sensationalist hype guy to boring practical guy ha! My impression is that he has been in no way been more qualified to host this podcast than the average layperson

@jtblau 2 месяца назад

I disagree. I would lump myself in with the average laypeople and I feel Kevin is way more qualified to host this podcast than me!

@lifebasesolutions3405 2 месяца назад

Does anyone else notice the edits? The change in voiceover? The second one is at 20:20. I am curious as to what you said the first time. Of course you fact check HatGPT. Right.