Latest / Never Post / AI At Work? In This Economy?

Transcript

AI At Work? In This Economy? · Never Post

0:10
Mike Rugnetta: Friends, hello, and welcome to Never Post, a
0:14
podcast for and about the Internet. I'm your host, Mike
0:16
Rugnetta, and we have a resistant show for you this
0:20
week. Hans talks with doctor Casey Fiesler, a professor of
0:25
information science at University of Colorado Boulder,
0:29
doctor Alex Hannah, director of research for the Distributed AI
0:33
Research Institute, and numerous workers from various industries
0:39
about being anti AI in the workplace, but being afraid to
0:46
say so or knowing that by saying so, you may be setting yourself
0:50
up for consequences because your boss has AI psychosis, because
0:56
of some vague pronouncement about the market, because your
1:00
coworkers are secretly in love with a chatbot. But hope is not
1:05
lost, Hans learns. It's just a little hard to see through the
1:09
thicket of slop.
1:11
First, though, we're gonna take a quick break. You're gonna
1:14
listen to some ads unless you're on the member feed. And when we
1:17
return, Hans, doctor Fiesler, doctor Hannah, and some of you,
1:21
yes, you, our listeners about being anti AI in the workplace.
1:42
Alex Hanna: Alright. I'm going to share screen. Hopefully, it
1:44
all goes apace, and nothing goes sideways.
1:50
Hans Buetow: It's May, and I'm in a webinar. And it's a really
1:53
good webinar. It's about the launch of the Luddite Lab. So
1:57
this is an online hub for organizing around issues of AI,
2:00
and it's run by the fine folks at DARE, the Distributed AI
2:05
Research Institute.
2:06
Alex Hanna: DARE or the Distributed AI Research
2:08
Institute is an independent research organization conducting
2:12
community rooted research as a globally distributed group of
2:16
academics, activists, and engineers who believe in
2:18
technology that benefits everyone.
2:21
Hans Buetow: I'm in this webinar because DARE is awesome. And
2:25
this project and this webinar, they've got the word Luddite in
2:29
the title, which is one of my self selected beats on Never
2:33
Post. So I'm here. I'm drinking tea. I'm snacking on snacks,
2:37
sesame sticks by the way, which my grandma called cat food and
2:41
fed me an awful lot of when I was little.
2:44
Suddenly, get a little slack ping from our host Mike Ragneta.
2:48
Mike Rugnetta: I wonder if there's a segment to be had in
2:52
looking at employees who are anti AI, but they're afraid to
2:57
say anything because of the position that their employer
3:01
takes.
3:03
Hans Buetow: I pause mid sip of tea, cat food just hanging out
3:06
of my mouth, because on my other screen, webinar chat is burning
3:10
up.
3:11
Speaker 4: This is extremely timely for me. My workplace and
3:16
academic library, yeah, we should know better, is going all
3:21
in on AI, and nobody will promise me that I'm allowed to
3:24
opt out.
3:26
Hans Buetow: There are a bunch of librarians in the chat
3:28
actually finding each other, finding other people who aren't
3:31
librarians.
3:32
Speaker 5: Wow. I work at a Fortune 500 company, and I
3:34
thought the AI Kool Aid was limited to the corporate world.
3:38
Very worrying that academia education is embracing this
3:43
madness.
3:45
Hans Buetow: The messages are collecting fast and furious
3:48
about people feeling alienated by AI in their workplace, and so
3:51
I spit out the cat food from my mouth. I slap on my Jordy
3:55
Laforte shades and my fingerless gloves and slam on the all caps.
4:00
This is exactly the webinar I'm in. Sip a tea, cat food. Oh, num
4:10
num.
4:12
Mike Rugnetta: Okay. So the answer is yes.
4:18
Hans Buetow: I sat there watching the chat some more, and
4:21
I suddenly had this weird sense memory of being at work during
4:24
the pandemic and watching the Slack chat during a staff
4:28
meeting while they were discussing masking or work from
4:31
home, and watching people furiously on main and then also
4:35
in side channels debating the biggest changes in our working
4:39
lives in probably a generation. And then I remember after a
4:45
while, those conversations just went away. Not that there
4:48
weren't people resistant to taking off masks or coming back
4:51
to the office, just that more and more each day, it seemed,
4:55
everyone was just doing it. The world was changing around us and
5:00
that shifted the context that turned our norms into
5:04
convictions. So, yes, lol.
5:09
It's too simple, I think to myself, as I swap out my Men in
5:13
Black Predator two Ray Bans. Sling on my Faraday gun bag and
5:17
crack my knuckles, accepting that I need to do a Never Post
5:21
story about this. Too simple. You know it's too simple. Mike
5:26
knows it's too simple.
5:28
Mike Rugnetta: I don't know. I'd wanna get some input from folks
5:31
who feel this way too.
5:33
Hans Buetow: Let's do it. Because in this episode, I want
5:35
to better understand the nuances of why it can feel just so
5:40
difficult to be an AI refuser at work. So, first step, do folks
5:46
actually feel uncomfortable about AI in their workplaces?
5:50
So, where to look? Easy.
5:54
Did you know we have a Never Post subreddit? R slash never
5:58
post. Do you have a story about a moment where AI was brought
6:07
into your workflow by a colleague, either mandated by
6:11
the organization or not, and you had to reckon with it. Turns out
6:21
people really wanna talk about AI at their workplace.
6:24
Jazzy Minotaur: Hey, Never Post.
6:25
Speaker 4: Hey, Never Post.
6:26
Mike Rugnetta: Hi, Never Post.
6:26
Talia: I saw the request on Reddit for awkward AI
6:28
interactions at work.
6:30
Hans Buetow: So I'm gonna put your responses into two
6:33
categories. The first was people watching their coworkers and
6:37
their customers normalizing the use of AI in a work setting.
6:40
Speaker 8: One of my coworkers has gone all in on it. I found
6:44
him putting in basic math problems into it instead of just
6:47
opening the calculator program. It has replaced most any
6:51
critical thought in his life.
6:53
Speaker 5: I was
6:53
Talia: asked to copyedit some text for a coworker. When I
6:56
returned it, they said, thanks, Chatchipitalia, which I hope and
7:01
believe, that they said with the intention of being polite, but
7:04
it most certainly made me feel deeply patronized or maybe like
7:09
objectified. Used beyond the regular scope of my labor for
7:13
sure. Deeply grossed me out.
7:16
Speaker 9: I work at an arcade and I have seen AI touching the
7:21
Jersey Jack pinball table based off of the Harry Potter license.
7:25
This table had its artwork done by an artist who uses AI tools
7:31
in their process. Public reception as I've talked with
7:34
other pinball players has been that they do not care about the
7:42
artwork being AI generated. All that matters is how the game
7:47
plays, and that is bizarre to me.
7:56
Hans Buetow: A lot of the responses in this category were
7:58
about people in parallel roles, so peers and colleagues. The
8:02
responses that involved supervisors or a whole
8:05
organization, those tended to fall into the other category of
8:08
responses, which felt a lot more existential.
8:14
Speaker 5: I work in health care IT. It feels wretched to have no
8:18
leverage to say no to something that's definitely gonna cause
8:21
potentially lethal problems down the road.
8:24
Speaker 10: For about ten years, was instructor in English,
8:27
college English courses specifically. When I pushed back
8:31
against kind of this AI embrace, they told me that they were
8:37
bored reading the same kind of wrote paper students wrote, so I
8:42
don't really care if they use AI. I mean, that's hard to,
8:46
like, hear. I think we should have pride in what we do as
8:48
educators and like, even if we get bored reading the same like
8:51
five business letters over and over again, that's our job.
8:55
Right?
8:56
Speaker 5: I work in reality TV and documentary post production.
9:00
I have a strong moral objection to using it myself, but when I'm
9:05
part of a team that decides to go down an ethically dubious
9:08
route as a majority or at the behest of a particularly AI
9:13
obsessed showrunner, it's a very uncomfortable feeling. I mean,
9:17
the nature of my role is such that, like, often the actual
9:21
generation of these elements falls on me. Hey. Can you create
9:25
a clone of Daniel's voice so we can have him say that he's doing
9:28
x for y reasons?
9:29
In the past six months, many network AI standards have become
9:33
markedly more lenient.
9:36
Speaker 11: In the last three or four months, IT leadership has
9:40
been reframing us from software developers to AI engineers. I
9:50
don't think I really have any fear of losing my job to AI at
9:54
this point, but I am very fairly confident that this AI engineer
10:01
path is going to be what the future of my career looks like.
10:05
It's been exhausting.
10:09
Hans Buetow: So with risk to reputation, risk to jobs, is it
10:15
really any wonder that some people don't feel the space to
10:19
be able to be viewed professionally as critical or
10:21
resistant to new technologies? I mean, precarity meet judgment. I
10:32
really started to understand what it meant to be wary of AI
10:35
while under pro AI scrutiny when I read a post from Jazzy
10:39
Minotaur who wrote, my VP of IT had a meeting where we could all
10:45
share our thoughts before major AI policy decisions got made. He
10:49
pretty much ignored every negative statement and acted
10:52
like he only heard supportive statements. Jazzy Minatar agreed
10:57
to talk with me, but only anonymously, which is, I wanna
11:01
say, notably, how most people responded.
11:06
Jazzy Minotaur: It was a meeting because our company didn't have
11:10
an AI policy. So anyone using AI was more or less violating
11:16
company policy. None of us wanted to be viewed as negative
11:21
on a technology. It makes us seem like the old guy who's
11:26
can't use an iPhone, you know, because they just aren't
11:29
adapting to new technology. So all of us wanted to seem like
11:32
there were a lot of positives with some negatives, But it
11:37
really became kind of clear that the negatives were just gonna
11:40
get ignored no matter what we brought up.
11:44
Hans Buetow: What is the risk to you if you get labeled that way?
11:48
Jazzy Minotaur: If you get labeled that way, then you are
11:51
probably viewed as only as efficient as you are the last
11:57
time you started updating. So you are only as useful as the
12:04
newest technology you've learned.
12:08
Hans Buetow: Look. I know that technology changes super fast
12:11
and that folks need to be certified. They need to be up to
12:13
speed. And I also know that the tech industry is funded by toxic
12:17
optimists. But if I were in charge, which I dare you
12:21
cowards.
12:23
But if I were in charge, I think I'd also value that perspective
12:26
that, hey. Maybe we should apply some critical rigor to this new
12:31
technology that seems to have some real risks and drawbacks.
12:36
But that's not how the tech industry works or at least how
12:39
the industry part of the tech industry works.
12:42
Jazzy Minotaur: I think there is a drive to bring costs down, and
12:48
the marketing is nothing but, hey, this will bring your costs
12:50
down. And any skepticism towards that is more just, kind of being
12:57
against the company doing better.
13:02
Hans Buetow: In the modern hierarchy of needs, it goes
13:04
number one, company, end of list. Hang on a sec. Have to put
13:11
on my monocle to read this. Next one. This fear of judgement?
13:17
This is everywhere in AI. Women are 25% less likely to use AI.
13:24
Why? Well, a recent paper from Harvard Business School asked
13:29
that question, and women cited this fear of judgment that if
13:32
they were to use AI, they'd be seen as dummies who need a
13:35
computer crutch to solve even basic problems. But their male
13:39
counterparts are seen as bold, as mavericks, as ingenious for
13:44
using AI.
13:46
And then there's the terror of your work being judged. AI
13:51
allows you to suddenly sidestep that fear. Have an LLM do it?
13:56
You have someone to blame when it goes tits up. This was summed
14:00
up for me by our listener Talia.
14:02
Talia: There seems to be a throughline of AI use at work
14:06
being about decreasing vulnerability, like minimizing
14:09
the attack surface. Terror at the judgment of their
14:13
professional outputs seems to drive people into the arms of AI
14:17
while predictably pummeling their capacity further. I want
14:23
my coworkers to know that their work is good, that their
14:27
thinking on a matter is valuable, that their labor is
14:31
deserving of respect.
14:38
Hans Buetow: The scope and power of this fear of judgment really
14:42
hit me when Jazzy Minatar described their attitude during
14:46
that meeting that they were having with their bosses.
14:48
Jazzy Minotaur: Most of us were trying to be balanced going in.
14:53
We, you know, have to look at them objectively even if we have
14:57
a moral objection to AI. We have to look at it from a company's
15:01
perspective.
15:03
Hans Buetow: At this point, I whipped off my specs, narrowed
15:06
my eyes, lit my desk on fire, and stared Jazzy Minotaur
15:10
directly in the black square that was their turned off video.
15:14
Those are two powerful words that are not gonna slip by me.
15:18
Why? Why do you need, if you have a moral objection to
15:21
something, to put that aside and look at it from the company's
15:25
perspective?
15:27
Jazzy Minotaur: At one point, it was expected that a person who
15:31
worked for a company would utilize their own ethics, their
15:35
Alex Hanna: own
15:35
Jazzy Minotaur: morals, and utilize those in decision
15:39
making. Today, though, we more or less are bowing to a
15:44
fiduciary responsibility to shareholders, and anything we
15:50
could do that would increase their value is more or less a
15:57
legal obligation for the company. When I'm acting as a
16:00
human, as a person, I'm allowed my morals. I'm allowed my
16:03
values, my ethics. When I'm acting as a company, I'm no
16:09
longer allowed those things.
16:14
Hans Buetow: I get it. I get it. I get it. But it also feels way
16:19
more bleak than I'm willing to accept. I can't really get over
16:24
this idea that companies are us, and we are companies.
16:29
It's like Citizens United, but, like, the other way around, and
16:33
that sucks too. So how do you be a moral person? How do you stand
16:39
against it, especially with management out there ready to
16:44
replace me with anyone else? Ideally, from their perspective,
16:49
with an AI agent.
16:53
Casey Fiesler: So obviously, social norms play a huge role
16:57
here. And I think that we don't yet have very strong social
17:04
norms around some of these things yet. And when we're in
17:07
this position where they're sort of in the process of being
17:10
formed, that's a really tricky place.
17:14
Hans Buetow: This is doctor Casey Fiesler.
17:16
Casey Fiesler: I'm doctor Casey Fiesler. I'm a professor of
17:20
information science at University of Colorado Boulder.
17:24
Hans Buetow: I needed Casey's help with those two words, moral
17:27
objection, and how they intersect with capitalism.
17:32
Casey Fiesler: The choice becomes between your morals and
17:38
being able to pay your rent. Olivia, it's a ton of pressure
17:43
because you might get fired if you don't use it, and that's the
17:46
kind of thing that absolutely can override moral objections.
17:50
Hans Buetow: I wanna be super clear. I get it. I've done it
17:53
often. And I will never judge someone for being forced to make
17:58
that bargain from inside of a system. Never.
18:02
But this is a heartbreaking reminder that surviving
18:06
capitalism is in a way its own moral obligation. I need to
18:10
live. You need to live. And since our being alive is so
18:15
dependent on our jobs, in our current system, it can come down
18:19
to you versus me.
18:22
Casey Fiesler: Suddenly, you're being pitted against other other
18:25
people.
18:26
Hans Buetow: AI is now a them in the middle of this fight, a new
18:30
combatant for the same jobs, and it's backed by these huge
18:33
corporations that are talked about as being completely
18:36
synonymous with the future of the entire US economy.
18:40
Casey Fiesler: So I do think that there's some push that
18:43
like, how do you get people to use your product? One of them
18:47
is, oh my god, you're gonna get left behind and you're, you
18:52
know, everyone else is gonna be like passing you, you're not
18:56
going be able to get a job, and you're hearing this from
19:01
students too, like I I I hear from students constantly who
19:05
tell me that they think that professors should never ban AI
19:09
because they all need to be using AI constantly because they
19:14
think they can't get a job if they aren't like good at using
19:19
AI and therefore should, you know, practice it at every
19:22
opportunity.
19:27
Hans Buetow: I had wanted to talk to Casey because of a video
19:30
that she had posted about this very thing, about student use of
19:33
AI.
19:34
Casey Fiesler: Because an interesting thing about the
19:36
proliferation of generative AI is that it hasn't just created
19:39
the ability to cheat, it has created incentives to cheat.
19:42
Hans Buetow: Incentives. Incentive to do something they
19:45
know is bad for them. To use a tool that according to a Brown
19:49
University survey of their students, around eighty eight
19:52
percent of undergraduates fear is affecting their learning and
19:55
cognitive capacity.
19:56
Casey Fiesler: People often use AI in learning context not
19:59
because they think that it's good for them, but because they
20:02
assume everyone else is using it. And if they don't, they will
20:05
be at a disadvantage.
20:08
Hans Buetow: Students start to assume that probably everybody
20:11
else is just already in these streets, just way ahead of them,
20:14
practicing AI every day. I don't know, whatever that means. The
20:17
fear of judgment is just rocket fueled by the mystery of who's
20:21
using it and who's not. Famously, saying to a young
20:26
person, hey. Everyone's doing it over and over and over again has
20:32
consequences.
20:33
Just this morning, I turned on the radio, and there was Bernie
20:37
Sanders yelling at me.
20:39
Bernie Sanders: We are already starting to see AI's impact.
20:44
There are estimates from very credible sources that tens of
20:49
millions of jobs could be lost here in The United States in the
20:54
next ten years. Whole professions wiped out and young
20:59
people finding it harder and harder to land entry level jobs.
21:06
Hans Buetow: Yeah. There are boogeymen everywhere. Look at
21:09
these influencers and companies, and they're all adapting the
21:12
tools. Look at over there. How many other people are going
21:15
after this job or scholarship?
21:17
Look at all these articles about how many students are cheating
21:21
with AI. Maybe maybe I'm the only one stewing about this, and
21:27
I need to cheat too. It's out of my hands.
21:31
Casey Fiesler: Regardless of how you might feel about AI morally
21:37
or self interested concerns about your own learning or
21:40
cognitive impact, you might use it anyway.
21:43
Hans Buetow: Suspicions are high that they are behind already. So
21:47
are we surprised that when you hold students jobs prospects
21:52
ransom, they're gonna do what you say? So what are workers, us
22:02
in any sector, what are we supposed to do against that same
22:06
onslaught of messaging? If you are someone who is against AI,
22:10
but you're feeling all this pressure and judgment, is there
22:15
another way to approach it?
22:19
Casey Fiesler: I do think that there is some level of knowledge
22:27
of AI and how it works that will help you use it better, like
22:34
more responsibly, and if you're the person at your workplace who
22:40
can explain to people why hallucinations happen, why this
22:47
needs to be double checked, and different ways that you might
22:51
think about using the technology that isn't going to get as much
22:56
backlash. There is a level of knowledge of AI especially in
23:00
terms of responsible AI use that could actually be of help in a
23:05
in a company, and if you're having, you know, moral concerns
23:09
about AI use, but there really is pressure for you to show some
23:17
level of competence in AI or some level of use of it, then
23:21
maybe that's the direction to think about. Can you be a voice
23:26
of reason? At least using AI responsibly if you're gonna have
23:31
to use it. So maybe that maybe that's a value add.
23:39
That's my anyway, if I if I had to have a little, glimmer of
23:43
hope, silver lining.
23:48
Hans Buetow: I appreciate Casey's optimism, trying to
23:50
balance moral obligation with a work obligation, and it's hard.
23:54
It's hard. But I wanna talk to somebody who's maybe trying to
23:59
do that. So as I say my goodbyes to Casey, I'm thinking, oh, boy.
24:05
That feels like the kind of person who would be willing to
24:08
use their name, who'd be willing to go on record, be vocal about
24:12
this with me.
24:13
Someone who, when confronted with an AI assignment from their
24:18
bosses, might say, I don't know, something like
24:20
Liana: You're gonna have employees who don't know how to
24:22
do x, y, or z anymore if you outsource it to some agent or
24:27
subscript. Right? Are you comfortable with that?
24:33
Hans Buetow: I gotta head to the van. I'm gonna grab my Velma
24:35
glasses for this one, so let's take a quick break. My job had
25:09
been to find someone who was owning their AI refusal at work,
25:13
doing what they can to develop and progress norms around AI in
25:17
their workplace. Amazingly, I found someone. Someone who
25:24
referred to themselves as well known as the AI hater in their
25:29
office, which, like Jazzy Minotaur, was a tech office.
25:33
Liana: I'm in the video game industry.
25:35
Hans Buetow: This is Liana who was unique among respondents
25:39
because she allowed me to use her actual factual name. Liana
25:43
is opposed to AI. And like Casey suggested, she's trying to help
25:48
the folks around her to better understand the tools that are
25:52
available to them.
25:54
Liana: I do my best to bring up the practical. I try my best to
25:57
make the quote unquote business case. Right?
26:00
Hans Buetow: So one day, Liana's in an all staff meeting, and the
26:05
company announces their new AI initiative.
26:09
Liana: And then almost immediately, because the chat
26:13
was not disabled, this isn't a Zoom call, the Zoom chat
26:17
explodes. If they're like artists, for example, they're
26:20
like, don't do this. The programmers are like, We're
26:24
like, we don't know what to use it for. And then there are
26:27
people, well, actually, truly great. Like, and there are
26:30
people, like, actually fighting each other, like
26:32
Hans Buetow: Wow.
26:33
Liana: In a work chat. No one is anonymous in this chat. You're
26:37
learning for the first time, hey, where do you stand on this?
26:41
And that was like, oh. That was hard to watch, actually.
26:46
Why? It almost I mean, it's that no religion, no politics, no AI.
26:54
Right?
26:57
Hans Buetow: Oh, yeah. This feels brutal, but honestly
27:01
correct. Your position on AI might be seen as a signal of the
27:07
type of person that you are.
27:08
Liana: I think it's hard to be a bystander. It doesn't feel that
27:12
way when you see how quick the chat's coming. You know what I
27:15
mean?
27:18
Hans Buetow: Once again, I'm immediately having flashbacks to
27:20
the dissolution of masking and remote work, where the shifting
27:24
context has made the norms into convictions.
27:30
Liana: You have to pick. Every time you do something at work,
27:33
you pick. Right?
27:35
Hans Buetow: Yeah.
27:36
Liana: Do I use this tool or not? Yeah. Every every decision
27:40
you make is what are you gonna use for this? You either write
27:46
your email with or without it. You either do your PowerPoint
27:51
with or without it.
27:53
I already came in with my bias. I'm like, no. I'm not. I'm not
27:58
touching any of this.
28:03
Hans Buetow: So Liana has been acting on her objections. She
28:06
tries to show colleagues how to do pivot tables, how to do
28:09
formulas in Excel instead of running AI. She copy edits her
28:12
own writing, and she's getting noticed for it.
28:15
Liana: It got written in my yearly review. Lists one to two
28:19
opportunities for the employee to develop in their role, And
28:23
then included in that section was they at utilization portion.
28:29
You should use this more. And I'm like, okay.
28:32
Hans Buetow: Yeah. How did that feel to to read slash talk to
28:36
your supervisor about?
28:37
Liana: I had a feeling it was going to happen.
28:41
Hans Buetow: Okay. Why?
28:42
Liana: Because we've already talked about it. One of our
28:46
employees who is relatively pro AI is like, hey, I found a way
28:51
to ask it about data in Excel sheet. And I'm like, you know
28:56
there's a formula for that, right? Like, you don't have to
28:59
use AI for this. And then on the side, my supervisor is, hey, we
29:05
can't stop people from using AI.
29:07
You shouldn't stop people or dissuade them from using AI.
29:14
Hans Buetow: What do you think it means that it got put in your
29:16
performance review and is a part of your growth opportunity? What
29:21
does that mean?
29:25
Liana: If I'm being pessimistic about it, it means that they can
29:32
write me up for it in the future is what I keep thinking about.
29:36
So I'm waiting in October to basically be like, hey, you
29:41
didn't do this. Like, it's a goal and then they're gonna say
29:44
I didn't meet it.
29:45
Hans Buetow: Why are you are you not meeting it?
29:48
Liana: No. I'm not meeting it.
29:50
Hans Buetow: I'm Are not you going to meet it? Are you like
29:51
purposely not meeting it?
29:52
Liana: I wanna see what they do.
29:55
Hans Buetow: Wow.
29:56
Liana: I'm going to see what happens if this is genuinely
30:00
something that they're tracking. Because as far as I know,
30:04
there's nothing tracking our usage.
30:12
Hans Buetow: Everyone I talked to said very clearly that work
30:15
has a different feeling from even a year ago. I mean, back in
30:19
those halcyon days of 2025 or, you know, dare we even remember
30:23
2024, generative AI was still enough of a novelty that
30:29
companies didn't need to have enterprise level solutions to
30:31
it. Its use was, at least for workers, early adopters who had
30:37
personal accounts, like a Claude account or a ChatGPT account,
30:40
and they would use them sometimes for work. So what has
30:44
changed in the last year? I think there's a case to be made
30:48
that in this past year, companies realized how much of
30:52
their proprietary information was being fed into these random
30:56
personal AI accounts of these employees.
30:59
No agreement, no policy, and definitely no oversight of where
31:04
that data was going. And I think it's likely that employers
31:08
started to get nervous. And I know that it's likely that the
31:12
AI companies who were all desperate for adoption, they
31:16
were more than happy to provide enterprise level solutions for
31:20
that nervousness. And then what I guess is that the c suite
31:25
suddenly had something to be bullish on. Right?
31:28
Despite them probably not fully understanding the why of it,
31:32
just the impulse for it. And you wrap all that up in these claims
31:37
of efficiency and cost despite AI being neither efficient nor
31:42
cheap. And now here, workers are at with each other. Some are
31:48
exhausted and resigned and wondering what the hell
31:51
happened.
31:52
Liana: I wonder if it would be easier if it was required. Like,
31:56
this is required. We are measuring you. Yeah. It you're
32:00
doing this whether we want to or not.
32:02
I wonder if that's I wonder if that's easier. I'm
32:04
Hans Buetow: not sure. In what way?
32:06
Liana: Because then you can sort of offload it and be like, I'm
32:09
only doing this because it pays my rent versus we still get to
32:15
pick and that gives you a lot of ownership still. It still feels
32:20
personal.
32:28
Hans Buetow: Personal ownership or paying rent. Is that really
32:33
the best duality available to the workers of the world? I
32:39
slowly took off my eye bobs at this point and rubbed my face. I
32:47
needed to get in touch with somebody who knew something
32:50
about gathering together, about working as a team, about
32:54
solidarity. Someone familiar, a voice in the wilderness.
33:00
Alex Hanna: My name is doctor Alex Hanna. I am director of
33:04
research for the Distributed AI Research Institute. I'm a
33:08
sociologist and writer on technology, labor, and politics.
33:14
Hans Buetow: And you ran the webinar that I that kicked all
33:18
this off for me.
33:20
Alex Hanna: Oh, no. Sorry about that.
33:21
Hans Buetow: Oh, it was a dull it's been a delightful journey,
33:25
and I have some questions about kinda where I've gotten to, if
33:29
that's okay.
33:30
Alex Hanna: Mhmm.
33:31
Hans Buetow: It was at this moment right here that
33:33
simultaneously both Alex and I whipped out our Breakfast at
33:37
Tiffany sunglasses, slammed them onto our respective perfectly
33:41
lit intensely focused visages. So if you're a worker
33:46
Alex Hanna: Mhmm.
33:47
Hans Buetow: How do folks who are in that worker class better
33:50
understand their own agency? Because it does feel like agency
33:55
is a word that's at the center of what we're we're talking
33:57
about here. Yeah. Because Dare has done so much work with folks
34:02
who are organizing organizing specifically around AI. How does
34:10
solidarity and organizing, which I know are not quite the same
34:13
thing Mhmm.
34:14
How do those intersect with AI being imposed at work so much
34:19
more?
34:20
Alex Hanna: Solidarity is important in a few different
34:23
registers. One of them is, one, being able to refuse and say no.
34:28
We're not going to agree that we have to use these technologies
34:32
and we're not competing we wouldn't have to compete against
34:35
each other. Hard to do that alone. And so the first thing is
34:41
refusal and being able to control where and when the
34:46
technologies are deployed.
34:47
So that's one element of it. Another element of it is cross
34:51
national solidarity. Maybe somebody is working with the
34:54
outputs of an LLM, but those outputs are not coming out of
35:00
the ether. There's a lot of guidance or input that has been
35:04
provided by data workers. Those data workers are often not in
35:08
the same workplace, they're either in a country that's far
35:12
away or they are folks that are receiving a fraction of what the
35:17
going wage is in a western company.
35:20
And so solidarity would look like building alliances across
35:25
different job classes. And so we've seen tech workers advocate
35:28
for data workers to folks in Kenya being paid $2 an hour. So
35:33
that's also very, very important element of that. And I'd say
35:37
that there's also a solidaristic element that comes from
35:41
understanding that the nature of one's work is more than just
35:46
one's tasks on paper. It's also like a cross racial, cross
35:50
gender solidarity.
35:52
I mean, it's that's something that's been a dynamic in US
35:56
labor at its best. It also kind of prevents this strike breaking
36:01
or this kind of scabbing that that we see in industrial
36:05
actions. I mean, I'd love to see solidarity with senior engineers
36:10
and junior engineers. I'd love to see job solidarity with
36:12
engineers and administrative assistants. Mhmm.
36:16
And that could change the nature of the relationship with
36:19
technology and ensuring that folks are understanding that
36:24
casualization might be happening in one area, but it could be
36:30
coming for them next. Actually, we're seeing a bit of a pretty
36:34
big wake up call too with a lot of people in the tech industry
36:38
who were told four years to learn how to code and now it's
36:42
learn how to use an LLM or you're going to be laid off. And
36:47
I think solidarity is a huge anecdote to that.
36:51
Hans Buetow: I want to put one more to you and see if it makes
36:53
sense that I've been musing on. Is it if part of the the weird
36:58
moment where we don't have these social norms yet between all of
37:01
the the ways that we're functioning because we're just
37:05
kind of like, I I am I the only one who's being left out? Am I
37:08
the only one not doing this? It seems to me that solidarity can
37:11
also help change the perception that everyone else is using it.
37:16
Change the perception that I'm gonna miss out.
37:19
Change the perception that I'm at risk if I don't. Because
37:22
suddenly you're able to talk, you're able to communicate, and
37:24
you're able to be like, wait, do you use it? Do you want to use
37:26
it? You can have those conversations in an open way in
37:29
a way that it doesn't feel like we have mechanisms or is even
37:32
really appropriate in a lot of these workplaces now.
37:35
Alex Hanna: Yeah. No. A 100%. I I completely agree with that. If
37:40
there are peer groups that are doing x y and z because there's
37:45
not really you don't really know what's happening, then it can
37:47
provide a little bit more insight.
37:49
I mean, that's one of the reasons we do our podcasts,
37:53
write our book is because it is showing that there is kind of
37:57
another way another way to act and be in relationship with
38:02
these technologies that doesn't require you to think that you
38:06
have to use them or that one must go on in the kind of
38:10
technological future that you don't have as much control over
38:14
as you'd like to or that could be actually empowering.
38:23
Hans Buetow: Alex, amazing. That's incredible. Thank you for
38:25
that. That's that's actually a really encouraging way to look
38:29
at it. Great.
38:32
I'm glad.
38:33
Alex Hanna: I'm glad.
38:39
Hans Buetow: Let's be clear. Forced adoption of AI is not
38:41
happening in every workplace, but it clearly is happening. And
38:45
when it's happening or it's even being explored, AI is changing
38:50
our workplaces and quickly. And it's way more complicated than
38:54
the machines will take our jobs. There's a moral and a practical
38:58
balance that has to be made on an individual level.
39:02
What does it mean to align your values and your actions around
39:06
this particular set of technologies in a work context.
39:12
There's also power to be found in communicating and working on
39:18
a collective level. So I think at the end of all this, despite
39:23
all my fears, despite my apprehensions, what I have is
39:28
two doctors, Alex and Casey, each independently telling me
39:33
that there are multiple ways forward, even when the boss says
39:40
there's only one. Thank you. Thank you.
39:55
Thank you. Thank you. You. Thank you. Thank you to everyone who
39:58
sent us a note, or a voice memo, or read something for us.
40:04
Thank you also to doctors Alex Hanna and Casey Fiegler. We're
40:08
gonna put links to both of their work in the show notes, but
40:10
we're also gonna link the Luddite Lab, which is the
40:14
resource that I was watching the webinar on. It's this amazing
40:17
thing, and it's made by DARE for organizing at work around issues
40:21
of AI. So if you are interested in hearing about how others have
40:25
done it, go and read the case studies and check out their
40:29
resources. It's really interesting.
40:31
It's really powerful and really useful. We are gonna keep
40:37
looking into AI in the workplace. So if you are a
40:40
person who is wrangling, was wrangling, or maybe has never
40:46
wrangled with AI in the workplace, send us, please, your
40:50
moral quandaries or if it's the case, your smooth moral
40:54
certitudes. Whatever you feel, we wanna know. Especially if
40:58
what you feel is that the definitions of what AI actually
41:03
does get misunderstood in your workplace.
41:09
You can find all the ways to get in touch with us in the show
41:12
notes.
41:52
Mike Rugnetta: That is the show we have for you this week. We
41:55
are gonna be back here in the main feed on or around Thursday,
41:59
August 13, the week after which is our member drive stream week,
42:04
during which we will unveil new merch, play new games, host new
42:08
guests, and hopefully add to our proud ranks a few new members
42:12
who get access to additional episodes, an ad free feed of the
42:15
show, and other fun things. Do you wanna become a member? Do
42:19
you want a deal, a steal, a reward for listening this far
42:23
into the episode? You can click the link in the show notes, the
42:26
one all the way down at the bottom.
42:28
Skip that first link. Skip all the other links all the way down
42:32
there. At the very end, you will find a link to a membership
42:34
discount that will run until stream week begins. You can
42:39
become a member for half off a yearly membership. That's right.
42:44
$24 for one year of Never Post. They're calling it the greatest
42:48
deal in podcasting. Become a member today at neverpo.st,
42:52
specifically. The link all the way at the bottom of the show
42:55
notes. And, you know, if you're listening to this later after
42:58
stream week, sorry.
43:00
You should listen to the end of the episode. Do know we put a
43:02
poem back here? Never Post's producers are Audrey Evans,
43:07
Georgia Hampton, and the mysterious doctor first name,
43:09
last name. Our senior producer is Hans Buto. Our executive
43:13
producer is Jason Oberholzer.
43:14
The show's host, that's me, is Mike Ragnetta. Of course, there
43:24
are victims in this poem. Victim. Victim. Victim.
43:28
Victim. Victim. Excerpt of History by Tracy k Smith. Never
44:05
Post is a production of Charts and Leisure, and it's
44:08
distributed by Radiotopia.