Aug. 4, 2026
Every Word You Said, Forever and Always
If knowledge I gave a company five, ten, or twenty years ago is still generating value for them, why does the payroll relationship get to end while the value extraction doesn't?
That's the question driving this episode. I read and respond to an excerpt from a LinkedIn post by a former colleague about AI systems that capture and index employee knowledge — the recordings, conversations, and stories that make expertise valuable — so all of it stays discoverable inside a company long after the person who generated it has left. Long-term extraction, in other words.
My issue isn't with the AI capability itself. It's with the assumption sitting underneath the post: that a company can capture, index, and reuse a person's stories, judgment, and point of view indefinitely, without renegotiating consent, ownership, or compensation as the relationship changes.
I trace it through consent, accountability, discoverability infrastructure, reputation risk, and the crux of it for me: compensation.
Key themes
- Consent has an expiry problem. Recording something while employed isn't the same as consenting to its indefinite reuse after departure.
- Accountability gets murky once context changes. If old material is repurposed years later and it backfires, who answers for it? The person who said it, or the company that recontextualized it?
- Discoverability is infrastructure. Making knowledge "searchable" means it gets indexed, pattern-matched, and interpreted, which raises the attribution question: am I cited as the source, and do I have any say in that?
- Reputation and digital footprint are at stake, especially if it's a company or context I no longer want to be associated with.
- The employment contract has changed. Call it the Hotel California Clause: you can leave the company any time you like, but your value never does. Payroll used to be the deal: compensation flows while you're actively contributing. Indefinite knowledge capture breaks that link. The value keeps flowing. The payroll doesn't.
- This isn't limited to employees. Contractors, freelancers, and consultants recorded and captured through AI face the same exposure, and it's nowhere in their contracts.
- Whose knowledge "counts" in the first place. I raise the open question of whether ageism, racism, or misogyny shapes whose expertise gets flagged as valuable enough to capture and monetize.
- Regulatory retention sits in tension with all of this. At my old employer in financial services, everything was already subject to audit and kept on record for six or seven years: recordings, emails, anything written. That's extraction for compliance rather than knowledge-mining, but it's still capture and retention I didn't fully control. I don't resolve that tension here. Where regulatory retention ends and unregulated knowledge-mining begins deserves its own conversation.
Reflection questions
Use these for journaling, discussion, or a follow-up episode:
- What consent does a company actually have to use my recorded conversations, stories, and expertise after I've left?
- If information I provided years ago is repurposed and the context has shifted, who is accountable when it backfires: me or the company?
- If a company builds infrastructure to make my captured knowledge discoverable, indexed, and pattern-matched, do I have any say in how I'm cited or attributed?
- If a former employer is one I no longer want to be associated with, what recourse do I have over how my past contributions continue to represent me?
- If my knowledge is valuable enough to keep generating value years after I've left, why doesn't the compensation relationship continue too?
- Has the employment contract changed, from "paid for your labor while employed" to "paid for your labor, but your knowledge stays extractable indefinitely"?
- Does this exposure extend beyond employees to contractors, freelancers, and consultants? Is it addressed anywhere in those agreements?
- Whose knowledge gets deemed valuable enough to capture in the first place? Is there bias (age, race, gender) in that determination?
- What would it take to make this explicit: legislation, individual negotiation, or structural change at the organizational level?
- As AI capture of video, writing, and audio becomes normalized, what belongs in a contract today that wasn't there before?
One-line takeaway
The post calls it preserving wisdom. I'm asking who owns that wisdom once the relationship ends, and why consent and compensation don't get renegotiated along the way.
This is an Assumption-Ground Audit
This whole episode is the Assumption-Ground Audit in action, pointed at a LinkedIn post instead of a policy. The AGA is the method I use to surface the assumption sitting underneath a decision before it calcifies into precedent. The question is never whether the technology is good or bad. The question is what everyone is treating as already-settled that hasn't actually been agreed to.
Here, the unexamined assumption is the Hotel California Clause itself: that indefinite capture and reuse of an employee's knowledge is simply part of the deal, and that consent, attribution, and compensation don't need renegotiating just because the mechanism changed from "someone remembers what you said" to "a system indexes it forever." That's the kind of assumption the AGA is built to catch before you build infrastructure, policy, or a vendor relationship on top of it.
If you're building AI systems that touch employee knowledge, client data, or institutional memory, those assumptions are hardening into your roadmap right now, whether anyone has examined them or not. If you want to know what they are before they're impossible to walk back, let's have a conversation.
1
00:00:24,741 --> 00:00:25,752
It's Nola Simon.
2
00:00:25,752 --> 00:00:32,741
I'm the host of the Hope and Possibilities
podcast, and I wanna talk a little bit
3
00:00:32,772 --> 00:00:41,421
about what's happening with reputation
and AI, and also communication.
4
00:00:42,211 --> 00:00:42,651
So
5
00:00:44,718 --> 00:00:46,508
I've written about this a little bit.
6
00:00:46,568 --> 00:00:48,278
I actually before…
7
00:00:50,468 --> 00:00:55,207
It was like maybe mid-July, I had
somebody who actually critiqued
8
00:00:55,347 --> 00:00:56,798
an essay that I had written.
9
00:00:56,917 --> 00:01:00,357
I've been very tr- clear and transparent
that I've been experimenting with AI.
10
00:01:02,308 --> 00:01:05,588
A challenge that I have is I don't
necessarily see a whole lot of use for
11
00:01:05,608 --> 00:01:12,718
AI in the stuff that I actually work on
other than actually helping me elaborate
12
00:01:12,908 --> 00:01:16,628
and write longer and more detailed essays.
13
00:01:17,067 --> 00:01:21,308
So I've ex- been experimenting
with that because, I like writing.
14
00:01:21,608 --> 00:01:25,448
The parts that I don't like about writing
are really actually related to training.
15
00:01:25,448 --> 00:01:31,008
So the takeaway questions, the
reflections, and stuff that people
16
00:01:31,008 --> 00:01:35,028
add on to essays to help people
really think for themselves,
17
00:01:36,278 --> 00:01:37,868
I'm not that interested in that.
18
00:01:38,078 --> 00:01:44,928
And so AI was helping me with that kind
of stuff because I would often actually
19
00:01:44,928 --> 00:01:48,477
record a podcast and then use the
transcript to really build out the essay.
20
00:01:49,187 --> 00:01:51,587
So it was based on my voice.
21
00:01:51,617 --> 00:01:56,108
It was based on… It was all transcripts
from actually doing the podcast.
22
00:01:56,687 --> 00:02:01,087
And then I could weave it into actually
something a little bit more essay-like,
23
00:02:01,087 --> 00:02:05,177
and then have these takeaway questions
to help people think for themselves and
24
00:02:07,647 --> 00:02:09,187
do those reflections.
25
00:02:10,557 --> 00:02:12,207
I'm a role model.
26
00:02:12,837 --> 00:02:14,947
I'm not a trainer.
27
00:02:15,077 --> 00:02:16,027
I'm not a coach.
28
00:02:16,067 --> 00:02:16,477
I
29
00:02:18,807 --> 00:02:26,777
really leading by example, and that kind
of training-type exercise isn't my forte.
30
00:02:26,877 --> 00:02:30,387
It's not really something that
I interest- I'm interested in.
31
00:02:30,387 --> 00:02:33,167
But I do know that, some
people do like that.
32
00:02:34,277 --> 00:02:39,237
So when I've been reflecting on how I've
been using AI, it's not really necessary.
33
00:02:39,518 --> 00:02:46,138
And this person critiqued me because she
felt that I hadn't… I'd lost my voice.
34
00:02:46,317 --> 00:02:51,208
I hadn't been specific enough, and the
way that I was using it it sounded more
35
00:02:51,208 --> 00:02:52,758
like Claude than it sounded like me.
36
00:02:53,927 --> 00:02:57,677
I didn't agree because I had
spent a lot of time on that, and
37
00:02:57,677 --> 00:02:59,498
as I said, I start with my voice.
38
00:03:00,118 --> 00:03:03,958
But what I found when I was actually
working through everything she said
39
00:03:03,997 --> 00:03:09,097
and, you know, the emotions surrounding
getting feedback that you don't love
40
00:03:10,137 --> 00:03:12,177
is how indefensible it really is.
41
00:03:12,177 --> 00:03:13,567
It's not like I can't write.
42
00:03:14,685 --> 00:03:15,545
I definitely can.
43
00:03:16,005 --> 00:03:20,715
But it's just more of, am I
experimenting for the sake of
44
00:03:20,715 --> 00:03:26,445
experimenting, trying to see where it
fits, and is it really that useful?
45
00:03:27,065 --> 00:03:33,375
So in terms of like SEO and AEO
and, attracting people to a website
46
00:03:33,375 --> 00:03:37,335
and coming up in search, that's
absolutely a, a part for it.
47
00:03:37,395 --> 00:03:40,355
I'm still going to be doing
stuff like that because,
48
00:03:42,585 --> 00:03:46,575
again, I'm not that interested
in figuring out keywords and how
49
00:03:46,575 --> 00:03:47,995
search works on the back end.
50
00:03:48,395 --> 00:03:49,545
I know that I have to do it.
51
00:03:49,545 --> 00:03:53,025
I know that I have to have information
where people you know, if they're
52
00:03:53,025 --> 00:03:55,965
trying to locate me or and they're
trying to locate somebody who thinks
53
00:03:55,965 --> 00:03:59,785
like me, it-- the systems will find me.
54
00:04:00,145 --> 00:04:02,125
That's really the
discoverability of it all.
55
00:04:03,525 --> 00:04:09,765
But what's happened now is how it
really affects reputation, and is
56
00:04:09,765 --> 00:04:14,125
it really worthwhile taking the
risk of experimenting publicly?
57
00:04:14,445 --> 00:04:15,775
And I don't think it is.
58
00:04:16,485 --> 00:04:19,495
So last week, Substack introduced Pangram.
59
00:04:19,715 --> 00:04:25,455
So that means that it's a system that you
can actually scan, work into, and have
60
00:04:25,455 --> 00:04:31,405
it rate whether it's 100% human or some
percentage of AI assist or completely AI.
61
00:04:32,045 --> 00:04:36,105
It's unreliable like all AI detectors are.
62
00:04:36,215 --> 00:04:39,935
But some people really like
the idea of being able to
63
00:04:39,935 --> 00:04:41,755
certify themselves 100% human.
64
00:04:44,691 --> 00:04:45,281
I don't.
65
00:04:46,431 --> 00:04:48,961
I think it feels like surveillance.
66
00:04:49,011 --> 00:04:54,851
I think it feels like the only people
who really benefit this, from this are
67
00:04:55,181 --> 00:04:59,401
Substack because they're saying that
they've been able to do something.
68
00:04:59,821 --> 00:05:03,351
And Pangram, who obviously get a load,
a lot of pr- publicity out of it.
69
00:05:04,091 --> 00:05:08,321
Apparently they're up for a new funding
round, and that's a big part of what
70
00:05:08,351 --> 00:05:09,901
their funding is going to be based on.
71
00:05:09,911 --> 00:05:14,121
So I'm trying not to talk that much about
them anymore now that I've learned that.
72
00:05:14,561 --> 00:05:14,901
But
73
00:05:18,469 --> 00:05:23,259
How does it benefit you to have
your users tattletale each-- on each
74
00:05:23,259 --> 00:05:29,789
other and mark yourself stuff, mark
other people as less than human?
75
00:05:30,759 --> 00:05:33,229
I don't see how that benefits us at all.
76
00:05:34,259 --> 00:05:39,809
And also, like such a thing as malicious
compliance exists, so you can't run it
77
00:05:39,849 --> 00:05:43,619
on posts that are less than 100 words.
78
00:05:44,069 --> 00:05:46,239
You can't run it on
anything that has audio.
79
00:05:46,239 --> 00:05:48,229
You can't run it on
anything that has video.
80
00:05:48,719 --> 00:05:51,009
You can't run it if there's a paywall.
81
00:05:52,139 --> 00:05:55,919
So personally, that's what I'm going to
do, is any experimentation that I'm doing,
82
00:05:56,369 --> 00:05:58,419
I'm going to move that behind a paywall.
83
00:05:58,979 --> 00:06:03,119
And, if I make anything public, it's
going to be less than 100 words.
84
00:06:03,209 --> 00:06:05,609
It's gonna have audio and
video in it because I know the
85
00:06:05,609 --> 00:06:07,029
systems can't run that yet.
86
00:06:07,479 --> 00:06:09,939
Now, that will eventually change,
and that's not gonna protect me.
87
00:06:10,430 --> 00:06:13,579
But I think that paywall really
is actually going to do something.
88
00:06:14,419 --> 00:06:17,949
And that's the other part is
like, why am I experimenting?
89
00:06:17,949 --> 00:06:19,419
What am I hoping to learn from it?
90
00:06:19,459 --> 00:06:21,190
Am I doing it just because it's there?
91
00:06:22,470 --> 00:06:24,889
Am I being lazy, really?
92
00:06:25,069 --> 00:06:30,250
And then LinkedIn saw the success
or f-f, backlash or whatever they
93
00:06:30,329 --> 00:06:37,940
considered Substack to be, and they
added a, seems like AI slop button so
94
00:06:37,940 --> 00:06:43,849
that anybody in the world could happen
across your post, report it, and then
95
00:06:43,849 --> 00:06:45,240
we don't exactly know what happens.
96
00:06:46,959 --> 00:06:49,500
LinkedIn, as usual, is
less than forthright.
97
00:06:50,139 --> 00:06:52,890
But honestly, it's right there
in the language and the button
98
00:06:53,379 --> 00:06:57,139
seems like, so there's no proof.
99
00:06:57,650 --> 00:06:58,889
There's no test.
100
00:06:59,330 --> 00:07:03,620
There's just, this is the way I'm
feeling about what you've presented
101
00:07:03,620 --> 00:07:07,780
as your work, whether it's AI
or whether you actually wrote it
102
00:07:07,780 --> 00:07:09,370
yourself and I just don't like you.
103
00:07:10,370 --> 00:07:14,089
Anybody in the world could
really just hit that report and
104
00:07:14,089 --> 00:07:16,670
LinkedIn would do something.
105
00:07:18,770 --> 00:07:22,800
Does that seem like it's got, the
potential to be ripe for abuse?
106
00:07:23,570 --> 00:07:25,290
Does that concern anybody?
107
00:07:25,350 --> 00:07:30,830
Since we know that ageism is a thing,
misogyny is a thing, racism is a
108
00:07:30,830 --> 00:07:33,770
thing, does that concern anybody?
109
00:07:34,400 --> 00:07:37,729
It's not like the internet is
a very nice place these days.
110
00:07:39,349 --> 00:07:43,310
And that brings me to a new thing
that happened over the weekend.
111
00:07:43,320 --> 00:07:48,309
There's a whole bunch of influencers,
and somebody corrected me that the person
112
00:07:48,310 --> 00:07:50,080
they knew was not actually an influencer.
113
00:07:50,080 --> 00:07:51,149
She actually had a business.
114
00:07:51,579 --> 00:07:56,939
But because she posts on In- TikTok and
Instagram and has large followings there
115
00:07:56,950 --> 00:08:03,300
and also attended a brand deal by OpenAI,
she's functioning like an influencer.
116
00:08:03,330 --> 00:08:12,639
So you… I think that the line
between thought leadership and
117
00:08:14,159 --> 00:08:16,679
influencer is a very interesting line.
118
00:08:16,820 --> 00:08:20,350
And I have also, going back to
that misogyny thing, heard that the
119
00:08:20,350 --> 00:08:25,270
word influencer often gets negative
connotation because it's mostly
120
00:08:25,270 --> 00:08:28,130
women who are successful influencers.
121
00:08:28,540 --> 00:08:29,820
A lot of them anyway.
122
00:08:31,100 --> 00:08:34,140
Anyways, this person that I've
been following for a while
123
00:08:34,140 --> 00:08:37,780
because she's interesting, she
has some really good ideas.
124
00:08:37,790 --> 00:08:44,346
She's, she's- She's got something,
and I can't tell you that I like her
125
00:08:44,346 --> 00:08:47,056
because there's a lot of things that
she says that I don't agree with.
126
00:08:48,236 --> 00:08:52,916
But she's captivating
in lots of ways, right?
127
00:08:52,926 --> 00:09:00,546
So I look at her as somebody I could
possibly learn from, mainly because I
128
00:09:00,606 --> 00:09:02,406
don't a hundred percent agree with her.
129
00:09:02,406 --> 00:09:06,766
It's like, what is she doing
that's so successful, and
130
00:09:06,766 --> 00:09:09,466
why does it bother me, right?
131
00:09:09,466 --> 00:09:11,886
'Cause in lots of ways, what
she says really bothers me.
132
00:09:11,896 --> 00:09:18,516
She deals with a lot of negative comments
and a lot of homophobia in her comments,
133
00:09:19,566 --> 00:09:21,856
and it's sometimes off the chart.
134
00:09:21,916 --> 00:09:25,256
I don't know that I'd want to
be publicly posting on TikTok
135
00:09:25,276 --> 00:09:26,486
these days, I gotta tell you.
136
00:09:26,716 --> 00:09:27,536
I have an account.
137
00:09:27,576 --> 00:09:28,786
I don't really use it that much.
138
00:09:30,526 --> 00:09:33,296
Anyways, so this brand deal
kinda went off the charts.
139
00:09:33,346 --> 00:09:34,296
She didn't help herself.
140
00:09:34,296 --> 00:09:39,646
She posted a a tour of her room at
this farm that they were staying in.
141
00:09:39,656 --> 00:09:42,076
They weren't paid apparently,
but it was transportation
142
00:09:42,076 --> 00:09:43,226
and accommodations and food.
143
00:09:43,796 --> 00:09:50,986
And she was talking about AI and how
wonderful OpenAI is while running a bath.
144
00:09:51,486 --> 00:09:55,016
You can imagine that the people who are
concerned about water consumption in data
145
00:09:55,016 --> 00:09:57,096
centers tore her apart in the comments.
146
00:09:59,116 --> 00:10:01,466
But this is the concern that I have.
147
00:10:01,996 --> 00:10:04,516
So are there concerns about AI?
148
00:10:04,576 --> 00:10:06,276
Absolutely, there are, right?
149
00:10:06,526 --> 00:10:08,086
There's environmental concerns.
150
00:10:08,086 --> 00:10:09,616
There's concerns about jobs.
151
00:10:10,006 --> 00:10:11,086
There's concerns about
152
00:10:13,356 --> 00:10:14,076
ethics.
153
00:10:15,016 --> 00:10:19,996
Like the, the list of concerns is so deep,
I don't even wanna go into it right now.
154
00:10:20,696 --> 00:10:20,976
But
155
00:10:24,946 --> 00:10:29,936
You have to look at th-this kind of thing,
and she's She posted again later, and
156
00:10:29,936 --> 00:10:34,516
she's "I didn't know this was gonna be so
controversial. I had no understanding that
157
00:10:36,546 --> 00:10:43,126
AI brought out this depth of feeling
in people." For one, how do you go
158
00:10:43,126 --> 00:10:48,306
on a brand trip for a company, and
the only thing you know about them is
159
00:10:48,306 --> 00:10:50,036
you think that you're a good brand?
160
00:10:51,566 --> 00:10:54,726
How do you not do the research,
especially when you are presenting
161
00:10:54,726 --> 00:11:00,966
yourself as a thought leader
who's capable of deep research?
162
00:11:01,386 --> 00:11:03,216
How do you not know
163
00:11:06,156 --> 00:11:08,566
anything about the reputation
of the company who's actually
164
00:11:08,586 --> 00:11:09,956
brought you on this brand deal?
165
00:11:10,886 --> 00:11:12,136
And she wasn't the only one.
166
00:11:12,176 --> 00:11:16,026
There were other people who were
invited to this thing as well.
167
00:11:16,536 --> 00:11:16,936
Now-
168
00:11:19,876 --> 00:11:23,186
It comes down to trust
and reputation, right?
169
00:11:23,586 --> 00:11:27,786
So do they deserve to be
torn apart in comments?
170
00:11:27,986 --> 00:11:32,266
No, I think we should be capable of having
like civil conversations presenting these
171
00:11:32,266 --> 00:11:38,276
concerns without name-calling, without
insults, and that's on both sides.
172
00:11:38,706 --> 00:11:41,886
It's not Chatham Rules
on the internet, right?
173
00:11:41,996 --> 00:11:44,726
And the internet is gonna do
what the internet is gonna do.
174
00:11:44,736 --> 00:11:45,866
Nobody can control that.
175
00:11:46,126 --> 00:11:51,096
So how do you present yourself in a
way that optimizes for trust while
176
00:11:51,096 --> 00:11:53,696
preserving your reputation, right?
177
00:11:53,726 --> 00:11:55,686
And it's not by calling people lunatics.
178
00:11:56,116 --> 00:11:57,896
It's not by being defensive.
179
00:11:58,526 --> 00:12:03,146
It's having a whole PR strategy if
you're going to be a public persona
180
00:12:03,146 --> 00:12:07,266
and you're actually going to be doing
brand deals and talking about things
181
00:12:07,266 --> 00:12:12,416
that are controversial, is knowing
your stuff and having that strategy
182
00:12:12,416 --> 00:12:16,526
of how you're owning it, how you're
explaining it, how you're promising
183
00:12:16,526 --> 00:12:20,386
it, and have those values that align.
184
00:12:20,726 --> 00:12:21,126
Now
185
00:12:25,014 --> 00:12:31,224
I think that becoming a thought leader,
becoming a creator these days is a
186
00:12:31,224 --> 00:12:36,194
really challenging position to be because
everything is shifting so much and values
187
00:12:36,194 --> 00:12:38,874
are really under pressure and optics.
188
00:12:39,284 --> 00:12:41,304
Everybody's examining the optics.
189
00:12:42,394 --> 00:12:49,104
It's beneficial to have a really good
understanding of what PR does, right?
190
00:12:49,154 --> 00:12:52,214
And how you preserve your reputation
191
00:12:56,448 --> 00:13:00,618
This is what's interesting to me
about the future of work because, it's
192
00:13:00,628 --> 00:13:07,238
not like platforms are not trying to
encourage people to be visible, right?
193
00:13:07,288 --> 00:13:09,328
LinkedIn just came out
with a collaboration thing.
194
00:13:09,338 --> 00:13:12,518
You can actually add like five
different collaborators to a post.
195
00:13:13,098 --> 00:13:19,178
It's designed to have people
talk about things, and that's
196
00:13:19,178 --> 00:13:20,298
the way the platforms work.
197
00:13:20,308 --> 00:13:23,158
The more you talk th- about things,
the more you get engagement.
198
00:13:24,108 --> 00:13:29,278
Then, ideally it's hopefully benefiting
you and your company, but also
199
00:13:29,308 --> 00:13:30,798
it's benefiting those platforms.
200
00:13:30,908 --> 00:13:36,678
So you really have to assess when you're
doing any of these types of activities
201
00:13:36,678 --> 00:13:38,748
like brand deals, like sponsorships,
202
00:13:41,058 --> 00:13:48,448
how is that company going to represent
you as your own personal brand, right?
203
00:13:48,478 --> 00:13:50,078
Do you want your name tied to them?
204
00:13:50,888 --> 00:13:53,768
I think that we have to
be a lot more careful.
205
00:13:53,768 --> 00:13:57,858
I think that it really comes down
to employees as well too, is do you
206
00:13:57,868 --> 00:14:06,518
wanna be associated long term with
these types of companies, right?
207
00:14:07,198 --> 00:14:11,028
I can't tell you that I would take
a brand deal like this, and honestly
208
00:14:11,028 --> 00:14:16,778
I've done previous podcasts about,
how I self-fund my podcast rather
209
00:14:16,778 --> 00:14:20,618
than look for sponsors because I
have concerns like this in terms of
210
00:14:22,866 --> 00:14:28,816
Reputation, and also what it means when
you're advocating for a third party that
211
00:14:28,936 --> 00:14:33,096
you have no control over, and you can't
really influence either in terms of their
212
00:14:33,096 --> 00:14:35,046
decisions and how they present themselves.
213
00:14:35,776 --> 00:14:38,496
It's the Wild West right now.
214
00:14:39,356 --> 00:14:42,276
I know that a lot of- … creators
make a lot of money from sponsorship
215
00:14:42,276 --> 00:14:44,836
deals- … and, brand partnerships
and stuff like that too.
216
00:14:44,836 --> 00:14:49,936
But I… There's a real concern
about trust with this on both sides.
217
00:14:51,176 --> 00:14:55,006
How do you manage that
reputation piece, right?
218
00:14:55,876 --> 00:15:01,146
And it's not just about AI, but AI
is infiltrating everywhere, right?
219
00:15:01,156 --> 00:15:04,946
So are you transparent
with how you're using it?
220
00:15:05,196 --> 00:15:09,646
Are you doing it in such a way
that it benefits you and your
221
00:15:09,676 --> 00:15:12,086
business- … and your relationships?
222
00:15:12,856 --> 00:15:19,396
And what kind of future is it helping
you build towards, not just, immediately?
223
00:15:20,426 --> 00:15:23,726
That's really what we have to look
into in terms of the future of work.
224
00:15:24,066 --> 00:15:28,356
So I think there's a real concern
about alignment of values,
225
00:15:28,486 --> 00:15:32,036
communication, trust, and reputation.
226
00:15:32,636 --> 00:15:38,796
And it's not just for people who are
self-employed as freelancers or creators.
227
00:15:39,326 --> 00:15:46,156
It's for employees and how they
represent themselves as individuals,
228
00:15:46,156 --> 00:15:51,096
but also as part of the brand, the
company that they work for, right?
229
00:15:51,106 --> 00:15:56,566
That's a big part of what your future is,
and, y- you're presenting them as part of
230
00:15:56,566 --> 00:15:58,666
your resume or part of your work history.
231
00:15:59,276 --> 00:16:04,096
It's important that you're aligned, and
that's something that I'm looking at.
232
00:16:04,376 --> 00:16:08,856
So how do you operate
safely in a way that-
233
00:16:10,948 --> 00:16:11,758
Feels ethical
234
00:16:13,868 --> 00:16:15,628
And is going to get you
where you want to be.
235
00:16:16,848 --> 00:16:18,768
Anyways, those are my
concerns for this week.
236
00:16:18,998 --> 00:16:19,958
Thank you for tuning in.
237
00:16:20,978 --> 00:16:22,058
I'd love to hear your thoughts
00:00:24,741 --> 00:00:25,752
It's Nola Simon.
2
00:00:25,752 --> 00:00:32,741
I'm the host of the Hope and Possibilities
podcast, and I wanna talk a little bit
3
00:00:32,772 --> 00:00:41,421
about what's happening with reputation
and AI, and also communication.
4
00:00:42,211 --> 00:00:42,651
So
5
00:00:44,718 --> 00:00:46,508
I've written about this a little bit.
6
00:00:46,568 --> 00:00:48,278
I actually before…
7
00:00:50,468 --> 00:00:55,207
It was like maybe mid-July, I had
somebody who actually critiqued
8
00:00:55,347 --> 00:00:56,798
an essay that I had written.
9
00:00:56,917 --> 00:01:00,357
I've been very tr- clear and transparent
that I've been experimenting with AI.
10
00:01:02,308 --> 00:01:05,588
A challenge that I have is I don't
necessarily see a whole lot of use for
11
00:01:05,608 --> 00:01:12,718
AI in the stuff that I actually work on
other than actually helping me elaborate
12
00:01:12,908 --> 00:01:16,628
and write longer and more detailed essays.
13
00:01:17,067 --> 00:01:21,308
So I've ex- been experimenting
with that because, I like writing.
14
00:01:21,608 --> 00:01:25,448
The parts that I don't like about writing
are really actually related to training.
15
00:01:25,448 --> 00:01:31,008
So the takeaway questions, the
reflections, and stuff that people
16
00:01:31,008 --> 00:01:35,028
add on to essays to help people
really think for themselves,
17
00:01:36,278 --> 00:01:37,868
I'm not that interested in that.
18
00:01:38,078 --> 00:01:44,928
And so AI was helping me with that kind
of stuff because I would often actually
19
00:01:44,928 --> 00:01:48,477
record a podcast and then use the
transcript to really build out the essay.
20
00:01:49,187 --> 00:01:51,587
So it was based on my voice.
21
00:01:51,617 --> 00:01:56,108
It was based on… It was all transcripts
from actually doing the podcast.
22
00:01:56,687 --> 00:02:01,087
And then I could weave it into actually
something a little bit more essay-like,
23
00:02:01,087 --> 00:02:05,177
and then have these takeaway questions
to help people think for themselves and
24
00:02:07,647 --> 00:02:09,187
do those reflections.
25
00:02:10,557 --> 00:02:12,207
I'm a role model.
26
00:02:12,837 --> 00:02:14,947
I'm not a trainer.
27
00:02:15,077 --> 00:02:16,027
I'm not a coach.
28
00:02:16,067 --> 00:02:16,477
I
29
00:02:18,807 --> 00:02:26,777
really leading by example, and that kind
of training-type exercise isn't my forte.
30
00:02:26,877 --> 00:02:30,387
It's not really something that
I interest- I'm interested in.
31
00:02:30,387 --> 00:02:33,167
But I do know that, some
people do like that.
32
00:02:34,277 --> 00:02:39,237
So when I've been reflecting on how I've
been using AI, it's not really necessary.
33
00:02:39,518 --> 00:02:46,138
And this person critiqued me because she
felt that I hadn't… I'd lost my voice.
34
00:02:46,317 --> 00:02:51,208
I hadn't been specific enough, and the
way that I was using it it sounded more
35
00:02:51,208 --> 00:02:52,758
like Claude than it sounded like me.
36
00:02:53,927 --> 00:02:57,677
I didn't agree because I had
spent a lot of time on that, and
37
00:02:57,677 --> 00:02:59,498
as I said, I start with my voice.
38
00:03:00,118 --> 00:03:03,958
But what I found when I was actually
working through everything she said
39
00:03:03,997 --> 00:03:09,097
and, you know, the emotions surrounding
getting feedback that you don't love
40
00:03:10,137 --> 00:03:12,177
is how indefensible it really is.
41
00:03:12,177 --> 00:03:13,567
It's not like I can't write.
42
00:03:14,685 --> 00:03:15,545
I definitely can.
43
00:03:16,005 --> 00:03:20,715
But it's just more of, am I
experimenting for the sake of
44
00:03:20,715 --> 00:03:26,445
experimenting, trying to see where it
fits, and is it really that useful?
45
00:03:27,065 --> 00:03:33,375
So in terms of like SEO and AEO
and, attracting people to a website
46
00:03:33,375 --> 00:03:37,335
and coming up in search, that's
absolutely a, a part for it.
47
00:03:37,395 --> 00:03:40,355
I'm still going to be doing
stuff like that because,
48
00:03:42,585 --> 00:03:46,575
again, I'm not that interested
in figuring out keywords and how
49
00:03:46,575 --> 00:03:47,995
search works on the back end.
50
00:03:48,395 --> 00:03:49,545
I know that I have to do it.
51
00:03:49,545 --> 00:03:53,025
I know that I have to have information
where people you know, if they're
52
00:03:53,025 --> 00:03:55,965
trying to locate me or and they're
trying to locate somebody who thinks
53
00:03:55,965 --> 00:03:59,785
like me, it-- the systems will find me.
54
00:04:00,145 --> 00:04:02,125
That's really the
discoverability of it all.
55
00:04:03,525 --> 00:04:09,765
But what's happened now is how it
really affects reputation, and is
56
00:04:09,765 --> 00:04:14,125
it really worthwhile taking the
risk of experimenting publicly?
57
00:04:14,445 --> 00:04:15,775
And I don't think it is.
58
00:04:16,485 --> 00:04:19,495
So last week, Substack introduced Pangram.
59
00:04:19,715 --> 00:04:25,455
So that means that it's a system that you
can actually scan, work into, and have
60
00:04:25,455 --> 00:04:31,405
it rate whether it's 100% human or some
percentage of AI assist or completely AI.
61
00:04:32,045 --> 00:04:36,105
It's unreliable like all AI detectors are.
62
00:04:36,215 --> 00:04:39,935
But some people really like
the idea of being able to
63
00:04:39,935 --> 00:04:41,755
certify themselves 100% human.
64
00:04:44,691 --> 00:04:45,281
I don't.
65
00:04:46,431 --> 00:04:48,961
I think it feels like surveillance.
66
00:04:49,011 --> 00:04:54,851
I think it feels like the only people
who really benefit this, from this are
67
00:04:55,181 --> 00:04:59,401
Substack because they're saying that
they've been able to do something.
68
00:04:59,821 --> 00:05:03,351
And Pangram, who obviously get a load,
a lot of pr- publicity out of it.
69
00:05:04,091 --> 00:05:08,321
Apparently they're up for a new funding
round, and that's a big part of what
70
00:05:08,351 --> 00:05:09,901
their funding is going to be based on.
71
00:05:09,911 --> 00:05:14,121
So I'm trying not to talk that much about
them anymore now that I've learned that.
72
00:05:14,561 --> 00:05:14,901
But
73
00:05:18,469 --> 00:05:23,259
How does it benefit you to have
your users tattletale each-- on each
74
00:05:23,259 --> 00:05:29,789
other and mark yourself stuff, mark
other people as less than human?
75
00:05:30,759 --> 00:05:33,229
I don't see how that benefits us at all.
76
00:05:34,259 --> 00:05:39,809
And also, like such a thing as malicious
compliance exists, so you can't run it
77
00:05:39,849 --> 00:05:43,619
on posts that are less than 100 words.
78
00:05:44,069 --> 00:05:46,239
You can't run it on
anything that has audio.
79
00:05:46,239 --> 00:05:48,229
You can't run it on
anything that has video.
80
00:05:48,719 --> 00:05:51,009
You can't run it if there's a paywall.
81
00:05:52,139 --> 00:05:55,919
So personally, that's what I'm going to
do, is any experimentation that I'm doing,
82
00:05:56,369 --> 00:05:58,419
I'm going to move that behind a paywall.
83
00:05:58,979 --> 00:06:03,119
And, if I make anything public, it's
going to be less than 100 words.
84
00:06:03,209 --> 00:06:05,609
It's gonna have audio and
video in it because I know the
85
00:06:05,609 --> 00:06:07,029
systems can't run that yet.
86
00:06:07,479 --> 00:06:09,939
Now, that will eventually change,
and that's not gonna protect me.
87
00:06:10,430 --> 00:06:13,579
But I think that paywall really
is actually going to do something.
88
00:06:14,419 --> 00:06:17,949
And that's the other part is
like, why am I experimenting?
89
00:06:17,949 --> 00:06:19,419
What am I hoping to learn from it?
90
00:06:19,459 --> 00:06:21,190
Am I doing it just because it's there?
91
00:06:22,470 --> 00:06:24,889
Am I being lazy, really?
92
00:06:25,069 --> 00:06:30,250
And then LinkedIn saw the success
or f-f, backlash or whatever they
93
00:06:30,329 --> 00:06:37,940
considered Substack to be, and they
added a, seems like AI slop button so
94
00:06:37,940 --> 00:06:43,849
that anybody in the world could happen
across your post, report it, and then
95
00:06:43,849 --> 00:06:45,240
we don't exactly know what happens.
96
00:06:46,959 --> 00:06:49,500
LinkedIn, as usual, is
less than forthright.
97
00:06:50,139 --> 00:06:52,890
But honestly, it's right there
in the language and the button
98
00:06:53,379 --> 00:06:57,139
seems like, so there's no proof.
99
00:06:57,650 --> 00:06:58,889
There's no test.
100
00:06:59,330 --> 00:07:03,620
There's just, this is the way I'm
feeling about what you've presented
101
00:07:03,620 --> 00:07:07,780
as your work, whether it's AI
or whether you actually wrote it
102
00:07:07,780 --> 00:07:09,370
yourself and I just don't like you.
103
00:07:10,370 --> 00:07:14,089
Anybody in the world could
really just hit that report and
104
00:07:14,089 --> 00:07:16,670
LinkedIn would do something.
105
00:07:18,770 --> 00:07:22,800
Does that seem like it's got, the
potential to be ripe for abuse?
106
00:07:23,570 --> 00:07:25,290
Does that concern anybody?
107
00:07:25,350 --> 00:07:30,830
Since we know that ageism is a thing,
misogyny is a thing, racism is a
108
00:07:30,830 --> 00:07:33,770
thing, does that concern anybody?
109
00:07:34,400 --> 00:07:37,729
It's not like the internet is
a very nice place these days.
110
00:07:39,349 --> 00:07:43,310
And that brings me to a new thing
that happened over the weekend.
111
00:07:43,320 --> 00:07:48,309
There's a whole bunch of influencers,
and somebody corrected me that the person
112
00:07:48,310 --> 00:07:50,080
they knew was not actually an influencer.
113
00:07:50,080 --> 00:07:51,149
She actually had a business.
114
00:07:51,579 --> 00:07:56,939
But because she posts on In- TikTok and
Instagram and has large followings there
115
00:07:56,950 --> 00:08:03,300
and also attended a brand deal by OpenAI,
she's functioning like an influencer.
116
00:08:03,330 --> 00:08:12,639
So you… I think that the line
between thought leadership and
117
00:08:14,159 --> 00:08:16,679
influencer is a very interesting line.
118
00:08:16,820 --> 00:08:20,350
And I have also, going back to
that misogyny thing, heard that the
119
00:08:20,350 --> 00:08:25,270
word influencer often gets negative
connotation because it's mostly
120
00:08:25,270 --> 00:08:28,130
women who are successful influencers.
121
00:08:28,540 --> 00:08:29,820
A lot of them anyway.
122
00:08:31,100 --> 00:08:34,140
Anyways, this person that I've
been following for a while
123
00:08:34,140 --> 00:08:37,780
because she's interesting, she
has some really good ideas.
124
00:08:37,790 --> 00:08:44,346
She's, she's- She's got something,
and I can't tell you that I like her
125
00:08:44,346 --> 00:08:47,056
because there's a lot of things that
she says that I don't agree with.
126
00:08:48,236 --> 00:08:52,916
But she's captivating
in lots of ways, right?
127
00:08:52,926 --> 00:09:00,546
So I look at her as somebody I could
possibly learn from, mainly because I
128
00:09:00,606 --> 00:09:02,406
don't a hundred percent agree with her.
129
00:09:02,406 --> 00:09:06,766
It's like, what is she doing
that's so successful, and
130
00:09:06,766 --> 00:09:09,466
why does it bother me, right?
131
00:09:09,466 --> 00:09:11,886
'Cause in lots of ways, what
she says really bothers me.
132
00:09:11,896 --> 00:09:18,516
She deals with a lot of negative comments
and a lot of homophobia in her comments,
133
00:09:19,566 --> 00:09:21,856
and it's sometimes off the chart.
134
00:09:21,916 --> 00:09:25,256
I don't know that I'd want to
be publicly posting on TikTok
135
00:09:25,276 --> 00:09:26,486
these days, I gotta tell you.
136
00:09:26,716 --> 00:09:27,536
I have an account.
137
00:09:27,576 --> 00:09:28,786
I don't really use it that much.
138
00:09:30,526 --> 00:09:33,296
Anyways, so this brand deal
kinda went off the charts.
139
00:09:33,346 --> 00:09:34,296
She didn't help herself.
140
00:09:34,296 --> 00:09:39,646
She posted a a tour of her room at
this farm that they were staying in.
141
00:09:39,656 --> 00:09:42,076
They weren't paid apparently,
but it was transportation
142
00:09:42,076 --> 00:09:43,226
and accommodations and food.
143
00:09:43,796 --> 00:09:50,986
And she was talking about AI and how
wonderful OpenAI is while running a bath.
144
00:09:51,486 --> 00:09:55,016
You can imagine that the people who are
concerned about water consumption in data
145
00:09:55,016 --> 00:09:57,096
centers tore her apart in the comments.
146
00:09:59,116 --> 00:10:01,466
But this is the concern that I have.
147
00:10:01,996 --> 00:10:04,516
So are there concerns about AI?
148
00:10:04,576 --> 00:10:06,276
Absolutely, there are, right?
149
00:10:06,526 --> 00:10:08,086
There's environmental concerns.
150
00:10:08,086 --> 00:10:09,616
There's concerns about jobs.
151
00:10:10,006 --> 00:10:11,086
There's concerns about
152
00:10:13,356 --> 00:10:14,076
ethics.
153
00:10:15,016 --> 00:10:19,996
Like the, the list of concerns is so deep,
I don't even wanna go into it right now.
154
00:10:20,696 --> 00:10:20,976
But
155
00:10:24,946 --> 00:10:29,936
You have to look at th-this kind of thing,
and she's She posted again later, and
156
00:10:29,936 --> 00:10:34,516
she's "I didn't know this was gonna be so
controversial. I had no understanding that
157
00:10:36,546 --> 00:10:43,126
AI brought out this depth of feeling
in people." For one, how do you go
158
00:10:43,126 --> 00:10:48,306
on a brand trip for a company, and
the only thing you know about them is
159
00:10:48,306 --> 00:10:50,036
you think that you're a good brand?
160
00:10:51,566 --> 00:10:54,726
How do you not do the research,
especially when you are presenting
161
00:10:54,726 --> 00:11:00,966
yourself as a thought leader
who's capable of deep research?
162
00:11:01,386 --> 00:11:03,216
How do you not know
163
00:11:06,156 --> 00:11:08,566
anything about the reputation
of the company who's actually
164
00:11:08,586 --> 00:11:09,956
brought you on this brand deal?
165
00:11:10,886 --> 00:11:12,136
And she wasn't the only one.
166
00:11:12,176 --> 00:11:16,026
There were other people who were
invited to this thing as well.
167
00:11:16,536 --> 00:11:16,936
Now-
168
00:11:19,876 --> 00:11:23,186
It comes down to trust
and reputation, right?
169
00:11:23,586 --> 00:11:27,786
So do they deserve to be
torn apart in comments?
170
00:11:27,986 --> 00:11:32,266
No, I think we should be capable of having
like civil conversations presenting these
171
00:11:32,266 --> 00:11:38,276
concerns without name-calling, without
insults, and that's on both sides.
172
00:11:38,706 --> 00:11:41,886
It's not Chatham Rules
on the internet, right?
173
00:11:41,996 --> 00:11:44,726
And the internet is gonna do
what the internet is gonna do.
174
00:11:44,736 --> 00:11:45,866
Nobody can control that.
175
00:11:46,126 --> 00:11:51,096
So how do you present yourself in a
way that optimizes for trust while
176
00:11:51,096 --> 00:11:53,696
preserving your reputation, right?
177
00:11:53,726 --> 00:11:55,686
And it's not by calling people lunatics.
178
00:11:56,116 --> 00:11:57,896
It's not by being defensive.
179
00:11:58,526 --> 00:12:03,146
It's having a whole PR strategy if
you're going to be a public persona
180
00:12:03,146 --> 00:12:07,266
and you're actually going to be doing
brand deals and talking about things
181
00:12:07,266 --> 00:12:12,416
that are controversial, is knowing
your stuff and having that strategy
182
00:12:12,416 --> 00:12:16,526
of how you're owning it, how you're
explaining it, how you're promising
183
00:12:16,526 --> 00:12:20,386
it, and have those values that align.
184
00:12:20,726 --> 00:12:21,126
Now
185
00:12:25,014 --> 00:12:31,224
I think that becoming a thought leader,
becoming a creator these days is a
186
00:12:31,224 --> 00:12:36,194
really challenging position to be because
everything is shifting so much and values
187
00:12:36,194 --> 00:12:38,874
are really under pressure and optics.
188
00:12:39,284 --> 00:12:41,304
Everybody's examining the optics.
189
00:12:42,394 --> 00:12:49,104
It's beneficial to have a really good
understanding of what PR does, right?
190
00:12:49,154 --> 00:12:52,214
And how you preserve your reputation
191
00:12:56,448 --> 00:13:00,618
This is what's interesting to me
about the future of work because, it's
192
00:13:00,628 --> 00:13:07,238
not like platforms are not trying to
encourage people to be visible, right?
193
00:13:07,288 --> 00:13:09,328
LinkedIn just came out
with a collaboration thing.
194
00:13:09,338 --> 00:13:12,518
You can actually add like five
different collaborators to a post.
195
00:13:13,098 --> 00:13:19,178
It's designed to have people
talk about things, and that's
196
00:13:19,178 --> 00:13:20,298
the way the platforms work.
197
00:13:20,308 --> 00:13:23,158
The more you talk th- about things,
the more you get engagement.
198
00:13:24,108 --> 00:13:29,278
Then, ideally it's hopefully benefiting
you and your company, but also
199
00:13:29,308 --> 00:13:30,798
it's benefiting those platforms.
200
00:13:30,908 --> 00:13:36,678
So you really have to assess when you're
doing any of these types of activities
201
00:13:36,678 --> 00:13:38,748
like brand deals, like sponsorships,
202
00:13:41,058 --> 00:13:48,448
how is that company going to represent
you as your own personal brand, right?
203
00:13:48,478 --> 00:13:50,078
Do you want your name tied to them?
204
00:13:50,888 --> 00:13:53,768
I think that we have to
be a lot more careful.
205
00:13:53,768 --> 00:13:57,858
I think that it really comes down
to employees as well too, is do you
206
00:13:57,868 --> 00:14:06,518
wanna be associated long term with
these types of companies, right?
207
00:14:07,198 --> 00:14:11,028
I can't tell you that I would take
a brand deal like this, and honestly
208
00:14:11,028 --> 00:14:16,778
I've done previous podcasts about,
how I self-fund my podcast rather
209
00:14:16,778 --> 00:14:20,618
than look for sponsors because I
have concerns like this in terms of
210
00:14:22,866 --> 00:14:28,816
Reputation, and also what it means when
you're advocating for a third party that
211
00:14:28,936 --> 00:14:33,096
you have no control over, and you can't
really influence either in terms of their
212
00:14:33,096 --> 00:14:35,046
decisions and how they present themselves.
213
00:14:35,776 --> 00:14:38,496
It's the Wild West right now.
214
00:14:39,356 --> 00:14:42,276
I know that a lot of- … creators
make a lot of money from sponsorship
215
00:14:42,276 --> 00:14:44,836
deals- … and, brand partnerships
and stuff like that too.
216
00:14:44,836 --> 00:14:49,936
But I… There's a real concern
about trust with this on both sides.
217
00:14:51,176 --> 00:14:55,006
How do you manage that
reputation piece, right?
218
00:14:55,876 --> 00:15:01,146
And it's not just about AI, but AI
is infiltrating everywhere, right?
219
00:15:01,156 --> 00:15:04,946
So are you transparent
with how you're using it?
220
00:15:05,196 --> 00:15:09,646
Are you doing it in such a way
that it benefits you and your
221
00:15:09,676 --> 00:15:12,086
business- … and your relationships?
222
00:15:12,856 --> 00:15:19,396
And what kind of future is it helping
you build towards, not just, immediately?
223
00:15:20,426 --> 00:15:23,726
That's really what we have to look
into in terms of the future of work.
224
00:15:24,066 --> 00:15:28,356
So I think there's a real concern
about alignment of values,
225
00:15:28,486 --> 00:15:32,036
communication, trust, and reputation.
226
00:15:32,636 --> 00:15:38,796
And it's not just for people who are
self-employed as freelancers or creators.
227
00:15:39,326 --> 00:15:46,156
It's for employees and how they
represent themselves as individuals,
228
00:15:46,156 --> 00:15:51,096
but also as part of the brand, the
company that they work for, right?
229
00:15:51,106 --> 00:15:56,566
That's a big part of what your future is,
and, y- you're presenting them as part of
230
00:15:56,566 --> 00:15:58,666
your resume or part of your work history.
231
00:15:59,276 --> 00:16:04,096
It's important that you're aligned, and
that's something that I'm looking at.
232
00:16:04,376 --> 00:16:08,856
So how do you operate
safely in a way that-
233
00:16:10,948 --> 00:16:11,758
Feels ethical
234
00:16:13,868 --> 00:16:15,628
And is going to get you
where you want to be.
235
00:16:16,848 --> 00:16:18,768
Anyways, those are my
concerns for this week.
236
00:16:18,998 --> 00:16:19,958
Thank you for tuning in.
237
00:16:20,978 --> 00:16:22,058
I'd love to hear your thoughts
