WaltGPT

Blackman’s AI problem — and the human behind it … Howie screams deadline! … And is that you, Phil?

When Lindsey Toussaint, the mother of a six-year-old with autism, emailed her lawmakers to ask them to support additional funding for the Division of Developmental Disabilities, she was already pretty skeptical about politicians.

So she was pleasantly surprised when she received a thoughtful reply from Republican Rep. Walt Blackman that thanked her for sharing her story and included details that proved to her that he actually read it.

Toussaint wrote back, thanking Blackman for being the first lawmaker to reply, “and also for taking the time to actually read my message for its personal details, which is refreshing and revitalizes the spark of hope in politicians.”

---

Now… Depending on how plugged in you are to the nexus of Arizona politics and artificial intelligence, you might know where this story is headed…

Blackman is an avid user of AI tools like ChatGPT — he’s been busted by the press a few times recently for using it to create a podcast and to self-publish an autobiography that embellished facts about his political record and his military service.

Not to mention, there are an awful lot of emojis on his Twitter feed.

When we called Tuissant to tell her that ChatGPT — not Blackman — had written that thoughtful reply, she was livid.

“I sat down and thoughtfully wrote that message. I'm a single mom of an autistic son, and I work full time, and I have also a part-time job, and I'm managing all my son's therapies and also conducting his habilitative therapy for him, trying to teach him how to be independent — I'm doing all of that on top of then writing a thoughtful, planned, very dedicated message to my politicians. And that's what I get in return?” she said. “I'd rather you just not respond at all.”

At this point, we hadn’t spoken to Blackman about the email.

But we’ve known him since his first election in 2018, and we know that he has a traumatic brain injury from his service in the Army.

We told Toussaint about the injury, and her tone immediately changed.

“I wonder if that will be mentioned as a reason. Here I am continuing to TRY to find faith in politicians,” she wrote in a follow up email to us.

Here at the Agenda, we have not been shy about shaming politicians for using AI to do their work for them.

Paradoxically, as Arizona’s most AI-forward news organization, we also lean on AI to help bolster our work.

If you thought we weren’t going to enlist our art intern, ChatGPT, to create some weird AI art for today’s edition, you’re clearly new here.

And that gives us a pretty nuanced view of the role that our new AI overlords can and should play in human political discourse.

And Blackman’s case is nothing if not nuanced.

We’ve known Blackman uses ChatGPT for a while. It’s obvious if you know how to spot it — so obvious that it’s kind of a joke around the Capitol. When we were asking around for Blackman’s new phone number this week, for example, one source responded, “LMK if you get a chat bot.”

And that view of him as a chatbot masquerading as human was bolstered recently by some solid reporting from the Republic’s Ray Stern, who noted that Blackman has made a host of inaccurate claims about his past, including his military service.

An autobiography Blackman wrote, for example, included serious errors — and some since-deleted passages that were clearly written by AI, like “Let me know if you'd like to integrate this into a larger narrative about your leadership style or campaign philosophy — I can help with that, too!”

When we called Blackman and told him we wanted to talk about his use of AI, he hung up on us.

Then he called back, saying he didn’t want to hide or be embarrassed about using AI to do his job.

“I have a traumatic brain injury from combat, and it helps me get what's in my mind to the email so I can get it to the constituents,” he said.

ChatGPT is one of the many workarounds he’s found to cope with his disability, he said.

When he left the Army, they gave him a voice recorder so he could mitigate his memory loss by recording conversations and replaying them. Using ChatGPT to craft emails is just another tool like that, he argues.

“It helps me process the things that I want to say. Because if I do not do that, it's hard for me to find words. It's hard for me to put those words to paper. And if you have not noticed, sometimes it's hard for me to get words out on the (House) floor,” he said. “In order for me to do the job, I have to use the tools that are that are there for me.”

And that’s something that often gets lost in the debate about how we use AI.

Yes, AI can be a tool for the lazy to act lazy. But it can also be a force for equity and uplifting people of differing abilities.

Toussaint’s main beef with a lawmaker using ChatGPT was her concern that he’s not reading her message and putting his personal time into crafting that thoughtful response.

Blackman acknowledged that he doesn’t actually “read” emails from constituents — but not like you might think.

Because his brain injury makes it hard for him to maintain attention and comprehension while reading, he has a computer program that reads his emails aloud to him. He’s much better at listening. He says he “reads” the news the same way.

And even though he’s using ChatGPT to write responses to constituents, it probably still takes him longer than most lawmakers.

He said he never responds to an email that he hasn’t “read.” And responding often means researching the issue, talking to staff, dumping notes into ChatGPT, generating a response and editing it before sending, he said.

“I don't think lawmakers should use AI just as a blanket sort of cover to do their work for them,” he said. “You should use it responsibly.”

And he acknowledges he hasn’t always used it responsibly. There have been embarrassing blunders, like with his book, which he maintains are accidental and not intended to deceive people.

Still, he’s not going to stop using ChatGPT to communicate with constituents, he said, because it helps him overcome his disability and do his job.

“I have headaches every day — on a scale of one to ten, they’re between a seven and an eight. People don't know this stuff because these are things I don't tell people,” he said.

But he knows that by not talking about his injury — or why he uses ChatGPT to help deal with it — he’s hurting his own credibility.

So from now on, he says he’ll put a disclosure on his emails explaining that he uses ChatGPT to help him write them, and why he uses it.

Blackman’s got a pretty legitimate reason to lean on AI. Plenty of lawmakers are doing the same — he’s just one of the few who’ve had to own up to it.

And adding a disclosure may not fix the trust gap Blackman has created. But we’re living in an era where politicians and chatbots are increasingly hard to tell apart.

A little disclosure might be the best we can hope for.

Interesting fact: If you tell people you use artificial intelligence, they trust you less.

It doesn’t matter whether it’s a teacher in a classroom, an executive looking for investors, or any other run-of-the-mill professional task. Most people just have a blanket mistrust of AI, University of Arizona researchers found in a recent study.

“I personally find this technology amazing,” one of the researchers, Martin Reimann, said. “But it's not just about what AI can do, it's about how it impacts human relationships.”

That’s the nut we’re trying to crack with our new weekly newsletter, the A.I. Agenda.

Our tech guru and A.I. Agenda editor, Adi Jagannathan, helps people make sense of this huge thing happening before their eyes, without reverting to cynicism, jumping on the hype bandwagon or just ignoring it altogether.

The fact is, sooner or later, everybody is going to have to decide what they think about AI and how they want to deal with it.

And time is short. Private equity firms are buying up real estate to build the data centers that power AI. Companies are revamping how they operate so they can replace workers with AI. Congress is trying to take complete control over regulating AI, and big lawsuits are determining who can use the intellectual property needed to train AI.

Meanwhile, AI has steadily become a part of the daily lives of millions of people since ChatGPT became widely available in late 2022. Students use it for homework. Radiologists use it to scan X-rays. Police sketch artists are using AI to make their drawings more life-like.

Now that AI has spread so widely, we’re at a cultural moment when millions of people can talk about their experiences with it, rather than having to listen to jargon-filled explanations from techies.

Just a few weeks ago, a professor who teaches the history of science at Princeton, D. Graham Burnett, wrote in The New Yorker about what happened when his students engaged an AI tool in conversation and then wrote up their observations:

“Reading the results, on my living-room couch, turned out to be the most profound experience of my teaching career,” Burnett wrote. “I’m not sure how to describe it. In a basic way, I felt I was watching a new kind of creature being born, and also watching a generation come face to face with that birth: an encounter with something part sibling, part rival, part careless child-god, part mechanomorphic shadow — an alien familiar.”

If we’re all going to come face to face with a “careless child-god,” who is surrounded by over-the-top hype or unthinking pessimism, we’re going to need a better GPS to help us navigate the world of AI.

That’s where the A.I. Agenda comes in.

The hype and the hyperventilating

When we started talking about launching the A.I. Agenda last year, we were a little skeptical that there would be enough news to report.

Now, after publishing the A.I. Agenda for the past five months, we’re endlessly curious about what’s going to happen next.

And there’s always something happening next: Humanoid robots that can dance. Court systems using AI-generated avatars. Lawmakers GPTmailing constituents.

Like a lot of people, we had a vague fear of AI. It conjured up images of an impending apocalypse, or at least our own jobless futures.

Five months later, we’re still worried about AI. But our vague fears have been replaced by specific concerns.

We’re asking ourselves questions like: Who gets to regulate AI? Who’s making money off of all this? Is Silicon Valley paying enough attention to public concerns? Are the new data centers going to suck up all of Arizona’s water? Can this thing catch my typos better than spellcheck? (Still no, weirdly.)

Instead of a paralyzing lack of knowledge, we now see questions and problems we can wrap our heads around.

We also see knee-jerk reactions from AI-phobes on social media. Many of them are keyed into the risk AI poses to creative work, like writing and painting, which are legitimate concerns.

But mostly, we get the sense that a lot of them haven’t engaged intellectually with how AI works and what it means.

We engage with it every week. We talk through what’s happening with AI, how it affects Arizona, how it all works, what it means for society, and how we can convey it to readers in a conversational way.

That has led to stories about companies asking for job applicants, but barring humans from applying. The intense competition between the U.S. and China over microchips. Arizona’s history of letting companies experiment in the state’s “sandbox.” Even being able to understand what birds are saying!

Simply put, we’ve built a better GPS for you to navigate this brave new world.

Don’t get lost. Get informed.

All you have to do is click this button.

DEADLINE!: As state budgeteers finally return to work today, Capitol scribe Howie Fischer breaks down what the budget fights are all about. Lawmakers are considering another round of the slush fund budgeting that they’ve leaned on to get bipartisan budget support since Gov. Katie Hobbs came onto the scene. Senate Republicans love the idea, House Republicans don’t, and Democrats have serious questions about how much wiggle room they have in the budget to create those slush funds anyway, Fischer tells Mark Brodie on KJZZ’s “The Show.” And the clock is ticking to get a budget done before the government shuts down on July 1.

“You know, the deadline is approaching, and it's amazing how much a deadline (helps). As you and I know, being in journalism, the editor yelling ‘deadline,’ and suddenly everything clears up and we can get a story out,” Fischer explained.

Mixed messaging: College students and the Arizona Students' Association rallied at the Capitol Monday (while lawmakers were on break) to advocate for “education affordability, free speech rights, protections for LGBTQ+ students and access to campus resources,” the State Press’ Carsten Oyer writes.

This land ain’t your land 🎵 …: Native American tribes and conservationists scored a victory in court against the proposed SunZia energy transmission lines that would carry wind farm electricity from New Mexico to Arizona after a federal appeals court ruled a lower court shouldn’t have thrown out their case, the Associated Press reports. The case essentially argued that the U.S. Department of Interior didn’t properly consult with tribes whose land the transmission lines would traverse, and if they had, it might have resulted in a historic property designation for Arizona’s San Pedro Valley and blocked the lines.

Make America Energetic Again: President Donald Trump wants to “save the Cholla Coal Plant in Arizona,” but it already shut down and would cost Arizona Public Service somewhere around $2 billion to reopen, KJZZ’s Wayne Schutsky reports. Republican Corporation Commissioner Kevin Thompson argues that Trump and other Republicans calling to reopen it only “promote financially reckless solutions that would burden APS ratepayers with nearly $2 billion in additional costs.” Meanwhile, a company that planned to build a hydrogen pipeline from New Mexico to Arizona is now saying it will carry natural gas instead, which, unlike hydrogen, contributes to greenhouse gases. But the company seemingly gave little heads-up to the regulators and stakeholders, including on the Navajo Nation, where pipelines would be laid, per Capital and Main.

“To the extent I may have been informed, but don’t remember, it was only in passing,” Corporation Commissioner Nick Myers said of the project. “I would love to talk to someone just to be in the loop.”

Spy games: Gov. Katie Hobbs vetoed legislation that would have banned the Chinese government and state-owned enterprises from owning land in Arizona, 12News’ Kevin Reagan writes. Advocates argued it’s necessary to stop Chinese spies, but Hobbs said in her veto letter that the bill would be “ineffective at counter-espionage,” doesn’t do anything to protect military assets and could lead to arbitrary enforcement.

Experience needed: TSMC isn’t necessarily worried about the hit that Trump’s tariffs would have on the company, the CEO told stakeholders on Tuesday, but it is worried the tariffs will hurt the global economy and tamp down demand for TSMC’s chips. The CEO also complained about the shortage of experienced labor in Arizona, per the Wall Street Journal.

Experience not needed: The Chandler City Council is officially asking voters to rewrite their city charter to ensure that city council members can become mayors, the Republic’s Lauren De Young reports. City council members have long been allowed to serve two terms on the council followed by two terms as mayor, but after Mayor Kevin Hartke considered running for a fifth term in office (returning to the council after two terms there and two as mayor) people started looking more closely at the term limits provision in the charter and questioning if he and other former council members should have been allowed to serve as mayor at all.

Elect us to another term? Upgrade to a paid subscription today.

Don’t take it off: Mesa Community College failed to act on sexual harassment complaints about a teacher who later asked students to strip in class “as part of a required exercise,” per Republic reporters Helen Rummel and Robert Anglen. The school finally opened a formal Title IX complaint against the professor, who was put on leave last month, but only after the horror stories hit the newspapers.

"There were lingering touches on my knees or shoulders," student Brecklyn Hall told the Republic. "My experience isn't nearly as insane as others. But it was really hard being in the class with him.”

Mogollon murder mystery: The New York Times picked up the story of two students from Phoenix’s Arcadia High School who were found dead while camping in the Tonto National Forest last week. Police haven’t released any details about the case, other than it’s being treated as a homicide.

Surprise City Council Member Johnny V. Melton helped out his constituents facing an unwanted, well, surprise.

He later updated his followers that the suspect “had the decency to wear a banana hammock.”

We can’t help but wonder if Laveen legend “Phil the Thong Man” has moved.

But we’re happy to see at least one local government official prioritizing getting rid of cracks in the road.