Funding for To The Contrary provided by: This week on To The Contrary how some women saw improved odds by applying for jobs, pretending to be men.

And will the next U.N.

secretary general be a woman?

Hello, I'm Bonnie Erbé.

Welcome to To The Contrary, a discussion of news and social trend from a variety of perspectives.

Up first, applying for jobs.

A wave of women on LinkedI has gone viral by experimenting with their profiles and pretending to be men.

Some found that presenting themselves in more masculine terms led to far greater visibility to employers almost overnight.

The results stirred frustration among women looking for their next jobs.

LinkedIn denied the gender effects reach, but many users felt the system reflects the same biases found in workplaces offline.

Joining me this week are feminist activist Erin Matson, Metr Conservative Media co-owner K.J.

McKenzie, Democratic commentato Nailah Amaru and the Washington Examiner investigations editor Sarah Bedford.

So the open question to you, Erin Matson, is, despite the fac that LinkedIn has denied bias, is there the same bias, worse bias, less bias online in job application than there is in the real world?

Well, Bonnie, unfortunately, bias against women in the workplace exists everywhere, whether it's online or off, and it's no surprise to see that it's being reflected on LinkedIn.

Well, first, this isn't research.

It's a single person anecdote with no control group, no peer review, no independent verification and no way to isolate variables.

So that's not data.

It's a personal blog post dressed up as science.

And I would say that even though it's a personal blog post, that there is evidence enough to write a blog post, and clearly that blog post is resonating with many other women, which indicate that there are other systemic issues that are not limite to just one personal experience, its an experience that is shared by many women.

One of the major concerns abou the widespread adoption of AI, which LinkedIn uses to filter these profiles, is bias.

Whether that's bias against conservatives, bias against women, racial bias.

It's a real problem that these biases are baked into the tools that are dictating things, like whether you appear to job recruiters.

But if biases and to—any one of you jump in on this— but if biases are baked into the tools, then how do we expect the online, you know, bias or lack thereof to be any better than it is in real life?

Well, Bonnie, I'll tell you one thing.

We can't count on the companies to sort it out for us.

And there's a reason for it.

It's—theres—very profit motivated.

The algorithm is their secret sauce.

That is actually where they'r deriving a lot of their private, their profit.

So they want to keep that information private.

They don't want to expose.

That is exactly why LinkedIn has been very cagey in its response about what is happening, because they don' want to reveal how this works.

So in order to really root that out, I think that we're going to hav to look at a number of things, whether it's more transparency required from governments.

That could be something that's considered or looking at impact, and that could be something that could be considered from neutral bodies.

But I also think another problem when we're talking about women is just simply the number of women working in this field is a huge issue.

And so when it's being built in a male dominated field and taking advantage of preexisting bias patterns, I think we're going to keep seeing these problems over and over again.

This is what I do for a living.

I study algorithms across all social media.

Algorithms don't think—they optimize.

And in doing so, they can amplify existing human behavior not invent bias out of thin air.

So they're built to reward engagement.

Clicks, comments, shares.

Watch time.

They don't car about race, gender or ideology.

They care about what people respond to.

So if certain content performs better, the algorithm shows it more.

Period.

How do we even out that?

I mean, that sounds like it's something you can never get rid of.

If society already has preferences and trends or cultural norms, algorithms can magnify them at a scale.

Not because it's biased, but because they're efficient mirrors.

The mistake the media makes, especially outlets like the Washington Post, is treating algorithms like ideological actors instead of tools.

When outcomes are uncomfortable, they blame bias instead of asking harder questions about content quality, audience behavior, or incentives.

This lady changed her whole makeup.

It wasn't about gender.

She changed her writing style.

She charged her headlines.

She changed tone, framing, messaging.

So those are major algorithmic variables.

LinkedIns algorithm boosts engagement, driving content, not chromosomes.

So you can't claim gender bias when you also rewired the entire content strategy.

That's what she did.

Well, I would—I would push back because I think that the algorithm rewards that bias that is built into the code because it's coded by human beings, that which is, predominantly, men.

And so when you have coding that is implicitly coded as masculine, changing whether that's something as explicit as changing you pronouns on the LinkedIn profile or changing job responsibilities from coordinated or collaborated to, drove and served as an architect, which are clearly rooted in gender coded language and the algorithm rewards the gender coded language with more visibility, with more engagement.

I think that speak to not necessarily the algorithm having bias by itself, bu by the process of its creation.

Building into the code the bias of human beings, which is then becoming tools which are now being amplified to again I think just kind of at scale, replicat the really problematic behaviors of what is perceived as leadership or masculine and how that's rewarded and how it's not depending on how people present, through their language.

And not just pronouns.

Through a more overall comprehensive profile.

I think two things can be true.

One is that it's just probably never been a better tim to be a woman in the workplace than in the year of our Lord, 2026.

Is it a perfect world?

No.

But women are not facing the types of systemic discrimination that previous generations had to contend with.

That being said, we're talking about a woman who went to an LLM, a large language model, and asked it to write copy that could be perceived a more masculine into her resume.

And that LLM is being trained on years and years worth of material that's already out there in the internet.

So you can talk about a code as written, but then it's training on millions and million and millions of pieces of data from across the internet that stretch back many years.

So it's going to reflect a lo of sort of institutional biases.

And that's hard to correct overnight.

And that's why it's so important the wa these LLMs are being trained.

But I think to KJ's point, you know, we're not talking about an experiment in which the women change their pronouns.

And then once they appeared as male in the algorithm, they were surfaced more.

We're talking about changing their profiles to have more direct language, more concise summarie of their skills and experience.

And that's something that when I was growing up and was taught resume writing that was just taught as what a good resume looked like.

It wasn't necessarily gender coded.

So a lot more about bias may be sort of grafted onto this narrative than is really there.

So are you saying that just tight writing, as we used to say in journalism school—direct, punchy, lively—is gender bias in some way?

That's what the article were discussing presented it as.

The ChatGPT model returned more concise copy in the titular woman's resume than she had in it before.

And that was the prompt she gave ChatGPT to produce that, was to produce for masculine copy.

But I think a really strong argument could be made that concise language in a resume isn't gender coded at all.

Exactly.

The gender coded, the gender language is this new age— I'm not gonna say it— but most platforms optimized for the same core signals engagement, retention, and shareability.

So in places like LinkedIn that means posts that are clear, confident, emotionally resonat and easy to react get rewarded.

So when everyone learns the rules of the game, people adopt that behavior to win.

Again, it's not censorship, it's incentive alignment.

But there is a trade off.

When creators optimize for the same algorithm, style converges before ideas.

So what you're seeing is the same tone, the same formats, the same buzzwords, the same emotional cues.

Now that's not becaus diversity of thought is banned, but because certain ways of expressing ideas perform better than others.

This doesn't eliminate dissent.

It penalizes complexity.

So again, nuanced slow or unpopular ideas struggle because they don't trigger instant engagement.

This is what it is.

So again these algorithms are content neutral.

But outcome driven is what it is Bonnie, I'm shaking my head because I have direct experience through my work professionally.

I lead an organization named Reproaction.

We work on abortion access issues.

We have had our content censored multiple times by META, content about abortion.

I have been in conversation with people at Facebook.

They notoriously hide the ball.

They disappear posts on a routine basis.

We've worked in coalition with colleagues who are having this issue.

And so some of the assumptions of what KJ is sharing are not accurate.

The algorithms are ultimately built by people.

This is not, this is not something that is just a computer calling the shots.

Although I also find the computers calling the shots very creepy.

That's a separate issue.

But there is, in fact, a heavy hand that is being applied in various companies.

And that was— I was speaking earlier about how their algorithms are their secret sauce.

They don't want to shar the details of how they do it, but it is absolutely the case that there is content that is being disfavored and pushed down actively by human beings, and we have had that struggle directly.

I want your thoughts on, since women are the majority, we're the majority population in this country.

Last time I checked, it was between 51 and 52% female.

Since women are makin or at least were pre Trump two, we're making all kinds of gains educationally and in the workplace in the professional fields.

What would you say to people who are assigned to go out there, make all the ChatGPT programs better in terms of reaching out to all kinds of applicants and bring in more of a representative group that's representing the loca population, while you're at it?

Yeah.

No, that's a great question.

I'd like to give a nuanced answer here.

I mean, I think to the point earlier about tight writing, I'm a writer myself.

Yes, we should be doing that.

We should be presenting our ideas forcefully and in a way that achieves results.

At the same time, the person who went through this on LinkedIn and had her post go viral, her blog post, there were terms that she used that are more gender coded.

And I think the fact that— well, give us some examples.

Like words like care, if the word care is a bad flag, I think that's the type of thing that we should also make sure that we're training ChatGPT, so it's not reinforcing bias against both language and traits that are traditionally ascribed to women because they're not worse.

Right?

And it's not this idea that caring is bad.

But we should ensure that that's also included in what is professional.

Sarah, what do you think about how to improve the algorithms and the results so that they seem to be more gender neutral?

Well, I think we're having two separate conversations here right now.

One is about ChatGPT, th LLM that the woman in question used to produce her copy for LinkedIn.

And the other is about LinkedIn's algorithm and what posts it choose to give more visibility versus less.

The algorithm question is one that conservatives have been wrestling with for a lot longer than I think liberals have, because conservatives were on the receiving end of quite a lot of bias over the past decade, from algorithms that social media companies that tended to reduce the visibility of posts that had conservative content in them.

When we talk about LLMs, absolutely, there's a responsibility of the people who are making these AI portals like ChatGPT or Grok or whicheve you're using to train the biases out of these so they can be useful tools for everyone.

The problem of algorithms in social media, that's on a sort of separate track about how those companies should be regulated, and what sort of content moderation powers they should be given for platforms that are often treated like public utilities or public goods.

And so those are sort of two separate conversations.

But both of them involve a lot of bias.

But do you think, let me just throw this out there to everyone, and then we need to switch topics.

But do you think all groups will ever be satisfied with any algorithm, or is everybody goin to feel ignored at some point?

Yeah, I think everyone, you know, it's going to be— people are going to pick with any and everything, n matter what the algorithm says.

But I think that we have to make the distinction between the algorithm and then companies actually censoring people, okay.

There are situations, as Aaron mentioned, where companies do censor base on graphic material and so on.

I mean she has a reproduction company.

If you're showing babies getting sucked out of you know, women, that's graphic.

If you'r if you're talking about abortion and things like that.

So that may be why some people are, you know, banned on YouTube.

I was, I actually was, censored on YouTube for something similar where I was talking about, you know, what abortio does for women and the process.

So I understand that.

But getting back to this language isn't male or female.

It's either effective or ineffective.

So what you guys are callin male coded is just competence.

So-called male coded words usually mean clear, assertive, results focused, confident.

That has—those types of of characteristics, of wording has been rewarded before algorithms were even here.

This is what people, this i what business owners want to see from employees, particularly in certain industries.

So those aren't male traits.

They're professional traits, and everyone benefits from using them.

I would just jump in and just say again, I, I definitely believe that, you know, there are, orientations t what is masculine and feminine, what is competent versus soft and how those show up in business language and how that shows up in overall you know, LinkedIn algorithms.

But I want to pivot back to the question that Bonnie asked in terms of, you know, well, you know, everyone will, you know, have something to say about no matter what the algorithm puts out.

I do think, you know, most concerning to m is the claiming of neutrality, right, that these algorithms are in fact neutral, as opposed to like the outcome, because I think that when platforms claim neutrality in how the algorithms are developed, it really does ignore the power of design choices, and how that impacts algorithms.

And again, going back to the human component of how these design choices are created, whether they bake in bias or have the opportunity, with some reflection and intentional design, to really expand access, and, you know, redefine merit and who has visibility and who doesnt.

Thats been a major conservative concern for years, actually.

All right.

Well, let us know what you think.

And whether you believe that all these algorithms should be followed up o and changed or just left alone, follow me on X @BonnieErbe.

From the workplace to the United Nations.

The UN has launched the hunt for a new secretary general to take over next year when Antonio Gutierrez, when his term ends.

Many observers expected the UN to choose it first female leader, who might well have focused on climat and international development.

But with Donald Trump's retur to the White House, many believe the search is shifting toward finding a pragmatic male with experience on security and power issues.

The UN Security Council makes its decision next year.

So, KJ, what do you see coming out of this decision?

Male, female or you don't care?

It doesn't matter to me.

I think just core attributes are what I'm looking for.

We need a leader who will tam down on nuclear proliferation, champions national sovereignty and calls out China's aggression.

Also defunds the UN's bloated bureaucracy.

Good luck on that one.

Even though we, there have been plenty of cuts in foreign aid in the US budget recently.

Still, it' very popular around the globe.

Your thoughts, Nailah.

If a woman is selected, it's certainly symbolic.

But it's also important to acknowledge that symbolism has its limitations.

But again, you know, if you have a woman who has the opportunity to change perception and change the priorities, you know, has the opportunity to be take seriously and kind of redefine what global leadership can look like in terms of reshaping those norms, which I think is important.

But again, to be clear, symbolism without the institutional backing and support is going to be very, very limited in terms of having any type of significant, substantive impact.

Sarah, if we're looking at bias, are we always going to come up with bias against women?

Well, if you're talking about Donald Trump's bias against women, I think every point of data— No, forget about Donald Trump just for the moment.

But generally speaking, because women over the last or 5 decades have been the one, the ones whom most of the discussion has been around bias against women.

Will they be the ones who will always be targeted, or found to be, I wouldn't say victims of, but the subjects of bias.

Well, in the context of this specific job search for UN secretary general, the applicant pool tends to be former heads of stat or other people who have risen to really high leadership positions in countries.

And those people tend to be male.

So there's already sort o a selection bias in the process.

But you also mentioned that the candidate was expected to focus on climate change and international development, but now may be someone who is more associated with national security and national sovereignty.

That's not necessarily a reflection of bias.

Those are just the times we live in.

All people of all sorts of political parties, even Democrats in the US and beyond, are moving away from climate change as a penultimate goal.

And I think that's more reflective of the times we live in than any sort of bias.

Everybody agree that that's a— moving away from climate change as a topic that needs to be tackled is going to happen globally?

Whether it should or whether it is a different question.

Right?

Climate change is an urgent crisis.

We are seeing rapid shift in extreme weather, and we know that this is something that needs to be dealt with.

But unfortunately, we are in a global moment right now of extreme instability and warmongering that is happening among a variety of nations.

And so that is something that is going to have to be the chief focus of the United Nations going forward.

You know, we have our own president has gone in and seized the president of another country and brought him in to the Unite States, has talked about going after Mexico after Canada, after Greenland.

So there's a variety of things that are happening right now.

And then you have Russia and Ukraine.

We have the issues, as KJ brought up, there's issues with China.

And so this is a period of extreme instability.

All of the ingredients see to be there for the possibility of significant, sustained, brutal, bloody, fatal wars.

And I think that's actually what the United Nations is going to be consumed with.

They're even talking about if the United States invades Greenland, which hasn't been taken off the table, that NATO will automatically be dissolved.

And so we're in the middle of major tectonic shifts in the world order.

But more than that I think it seems highly possible that we could be headed towards some major wars.

Does everybody agree?

We're heading towards wars, or are the wars coming from the Trump administration, which was, of course, behind the move to grab Maduro which was unprecedented, really?

No, we are not.

I wholeheartedly disagree with everything Erin just said, and it' not even backed up with facts.

First of all, as some— if you have been watching the news, most Venezuelans are happy about the liberation and removement of this terrorist, this narco terrorist who has never been, been shown to be the president of Venezuela.

Not only that, but as someone who has family in Trinidad and Tobago who border Venezuela, they were even shouting in the streets, happy that this man was removed by Trump because Venezuelans we saw a third of the population have moved, just walked out due to the the inhumane treatment of his people.

This was a liberation of Venezuela.

What are the chances at this point?

Do we all agree at zero, 25%, 50%?

Something else that a woman will be appointed secretary general to the United Nations?

And why is it taken us so long to get to the point where we still haven't had one female UN secretary general?

Bonnie, it's heartbreaking, right?

I mean, we've never seen a woman as UN secretary general.

It is such an important position.

There are absolutely women who are qualified and able to serve, and I don't want to come from a place of despair.

One thing that I have foun irritating in the news coverage of this is this idea that, well, because Trump is president of the United States, we should just check out of the idea of putting a woman into that position that we should just proactively say, forget it.

We're not even going to try.

And so I, I will b one of the first people in line, crying tears of jo if we get someone like Michelle Bachele or another very qualified woman.

Those women are out there, they're available.

And if not this round, I hope we get it soon.

But in any case, whoever takes this role is going to have a real hot mess on their hands.

There's zero reaso why Donald Trump being president would somehow discourage the appointment of a woman to that position.

I mean, during his first term, we sent a woman to represent the US at the United Nations.

During his second term, he tried to send a woman there, Elise Stefanik.

The nomination didn't work out, but there's no evidenc that Donald Trump would somehow work less with the U if a woman was in charge of it.

But we're talking about a world that has been really unstable for the past few years, and has become much more stabl under Donald Trump.

In Ukraine, you've seen finally, after years of stalemate, progress toward negotiations toward the end of the war that Russia started, when it invaded Ukraine, you've seen hostilities wind down between Israel and the Palestinians in Gaza.

We'll have to wait and see what happens in Venezuela.

There's a real discussion to be had.

This is not what Trump supporters voted for, its no what America First looks like.

But if the United States engagement in Venezuela is limited or none at all, and Venezuela is able to come out of the shadow of a dictatorship, that would be another check in Donald Trump's sort of effort to make the world a more stable place.

So there are certainly global problems out there, but I would argue that almost none of them are of Donald Trump's making.

All right.

That's it for this edition of To The Contrary.

Keep the conversatio going on social media platforms Instagram, Facebook, X and TikTok.

Reach out to us @tothecontrary.

Visit our website, the address is on the screen and whether you agree or think to the contrary.

See you next time.

Funding for To The Contrary provided by: You're watching PBS.