With the above in mind, let me share a cautionary story about a CIA-occupied building and an iconic U.S. Senator, the late Abraham Ribicoff. I’ll warn you of the grubby details ahead, but please stick with me. The same hair-pulling scenario could have played out at the local or state level.

Jon and I both scoured hundreds of federal real estate leases in search of the names of politicians and their buddies. I actually had it worse than Jon, who, for plot purposes, caught up with the leases much more promptly.

Even with help from one of Senator Edward Kennedy’s committee staffers, I struggled for weeks and weeks to get the leases under the Freedom of Information Act.

Granted access, I then became an honorary bureaucrat. Each day, I showed up at the General Services Administration headquarters at 18th and F Streets N.W. for the latest batch.

That still didn’t do the trick. For eons, bureaucrats had been ignoring the law and not listing the names of partners in countless partnerships.

Now flash ahead fifty years from the 1970s to 2024 (and beyond).

In a bot-friendly world, optimized for reporters using ChatGPT and its brethren, the leases would have revealed the names of all partners in partnerships. But the bureaucrats at the General Services Administration blithely flouted the law. Even if the GSA had been fully computerized back then, the ‘crats could have manipulated their databases to hide facts from me. Some unscrupulous GSA people might even have used their own “fixer” bots to thwart investigative reporters and law enforcement types.

One way to counter this would be a mix of smart bots and well-motivated humans with at least a modicum of canniness and emotional intelligence.

In Ribicoff’s case, I already knew from others’ reporting that he was hiding behind the name of a trustee named David Kotkin, his former law partner, in various Washington real estate investments. Now, what about the federally leased building at 1200 Wilson Boulevard in Arlington, Virginia, across the Potomac from D.C.?

Kotkin’s name was missing from the lease. I saw only the names of  the two general partners in the “Wilson Associates” partnership–not all of them. But Kotkin did show up in local real estate records. How could I nail down Ribicoff’s connection with the high-rise through Kotkin if the facts justified this? Here’s what happened, as I reported for the old States News Service in 1975:

“Asked by this reporter how he had heard of this investment opportunity, Kotkin said, ‘I don’t remember ever knowing anything about it. Why?’

“When told his name was on the document and asked whether he might have been representing Ribicoff, Kotkin—Ribicoff’s former law partner from Hartford—replied, ‘Oh yes, yes, that’s what it must be…’”

Could a bot have succeeded in prying the information out of the reluctant Kotkin? Imagine the embarrassment for both him and his ex-law partner. In the past Ribicoff had fooled the media into thinking he’d avoided investments in federally leased buildings. He sold his Wilson Associates stake in 1968, two weeks after winning re-election. Ribicoff sat on a committee overseeing GSA and was a close friend of Charles E. Smith, the federal contractor associated with the CIA-occupied high-rise. GSA is the government’s real estate arm, among other roles. The story made the NBC evening newscast, but without my getting Kotkin to say, “Yes, yes,” that would never have happened. Would current AI have sufficed?

Now—what about my travails getting the leases in the first place? Could a bot have successfully lobbied the Kennedy aide and others for a look at the leases?

I suppose AI agents could have bargained with each other, but don’t count on that scenario happening tomorrow. Oh, to ponder the possibilities for the future! Will a humanlike bot deal with another pseudo-human working for a future Senator Kennedy or his silicon equivalent? My hope is that human good guys can remain on the job in both newsrooms and the Capitol building—and use bots smarter than the bad guys’.

Last year, in a Substack essay titled Can AI Liberate Journalism, James O’Shea raised some related questions. He is a former editor of the Los Angeles Times, also served as managing editor of the Chicago Sun-Times, and earlier worked as a courthouse reporter for the Des Moines Register.

The essay reported O’Shea’s conversation with a North Carolina tech entrepreneur named James Boyd. At one time, O’Shea said, Boyd helped Lockheed Martin cook up an intricate computer-generated world for troops dealing with new “languages, cultures, and military threats.”

O’Shea wrote:

“Boyd readily acknowledges the limits of a computer-generated journalism model. How would a source, for example, react to a call, email or text from a computer asking it a question?

“’Would people respond to questions the same way as they would to a skilled reporter,’ he asks, ‘someone who is trying to build trust and ask questions of to get (a source) to open up the same way we see in movies or probably the way you experienced? I think that’s more difficult.’

“But, he cautions, machine learning and artificial intelligence systems enable computers to do many things that people didn’t think they could a few years ago, like play chess better than Garry Kasparov, the Russian world chess champion defeated in a six game match by IBM’s Deep Blue supercomputer in 1997.”

So keep an open mind about what might happen in time—a far different question from what should happen.

And this isn’t even to consider the myriad of other unknowns for journalists and nonjournalists alike as bots proliferate. Employment security, of course, is at the top of the list, especially with the eagerness of so many corporate executives to slash away jobs. Not to mention privacy worries, along with racial and ethnic stereotyping.

Meanwhile, I’ll immodestly recommend The Solomon Scandals for hedge funders, private equity people, and other Wall Streets eager to automate newspapers to the max.

So much of the book is about the nuts and bolts of investigative journalism. Scandals isn’t Moby-Dick—it’s a character-driven suspense novel, packed with ethical horrors written up in loving detail. But the same idea applies. You can’t fully understand whaling without knowing what “cutting-in” or “Nantucket sleigh ride” means.

Likewise, you’d do well to familiarize yourself with the details of serious journalism so you can better distinguish your media properties from the competition. Bot-izing your newsrooms at the expense of the real-life Stones is not the best idea. Use AI for formulaic work, not for higher-level journalism; and instead of replacing humans, give them a chance to work in partnership with bots to improve both productivity and quality. Are you listening, Gannett and McClatchy—and Wall Street?

As for independent journalists and other writers, take hope! For now, at least, many wonderful AI tools are online for free or at least at a low cost. I invite you and others to give the ScandalsBot a try for an imperfect but still useful look at the novel, including the issues that pop up in my chamber of journalistic outrages. No bots in the book itself. But who knows what’s ahead, ethically and otherwise, if we can’t intelligently extrapolate from past and present realities?

Image credit: Meta AI.