Just Making Things Up

Authenticity is key to creating a compelling story. Without it, the story falls flat -- or worse, the end user knows more than the author and can correct them. Oof.

The constant research into the area where FTS takes place is an effort to make the piece feel authentic. As I noted in the previous post, the area's accent is central to its success. Pandemics and politics have kept me from finishing this piece, but I will. I must.

I replayed my two-hour conversation with six local intellectuals. One topic that surfaced was how people are prone to making things up -- and how central locating the facts is to a project's accuracy and authenticity. One of the scholars noted that people often do not intentionally invent things; they simply mix up information with stories, memories, or legends.

He told the story of a person who was said to have studied medicine at the Mayo Clinic in Rochester, Michigan. The local newspaper was about to dedicate a story on his life. That is, until the facts were checked. The Mayo Clinic registrar did have a record of someone with that name enrolled in the medical school there, but at a different time -- twenty years later, in fact. Then, when a pack of envelopes was carefully pored over, invoices from the Rochester Institute of Technology showed that the person had actually studied radiology in New York State.

Discovery adds a layer of research -- and irony -- to my project, since many of the places I visited have been transformed over ten decades. The journey itself adds the layer of authenticity that completes the bridge.

Fact or Fiction? Real or Fake?

Creating generated imagery is challenging—but the challenge is ethical. Some fear being replaced by boorish hacks who appropriate others’ work; while this is possible, that concern ultimately exposes itself for what it is.

More pressing is how serious artists use tools—any tools—with ethical practice. Once, coloring images digitally seemed unthinkable compared to traditional methods. Yet digital work demands the same depth of knowledge and sensitivity. Without that understanding, the result is predictable and uninspired.

I gave a brief talk the other day on FTS at the college where I teach. I showed the 3-minute video with the spoken-word narration in my previous post. The room gasped when I told the audience that the man's voice was sampled from generated AI. A professor asked me about my ethical use of AI in this case, and I said I had no problems with it, since the voice was generated by a computer. I mentioned my pursuit of the comedian and spoken-word raconteur Reinaldo Ruiz and how a sample of his voice could more accurately express the color of the language in Cordoba. That statement generated even more gasps.

I counter that with:

• Music is routinely sampled. How is sampling a voice by AI any different?

• Is tweening in animation legitimate image processing?

• Decades ago, engineers programmed musical instruments in songs in all genres of music. 

• And CGI is responsible for...finish this sentence for me, please.

Is it the degree? The Origin? The Attribution? 

For my project, studying the language of someone like Reinaldo is essential. Achieving authenticity in the spoken work is where the artistry lies when using a new tool.

 

The Long Story Continues

 While many scorn AI, I think it's awesome. What most of my creative colleagues believe will happen to them, I think, won't. Technology serves us, and fearing it will just get in the way of using it creatively. In reality, AI can make work better....if you're already doing good work. Bad work will always weave its way in. People who made shitty images on film went on to make shitty images digitally. 

It's argued that the digital camera has ruined photography. But the same could be said for the hand-held camera; in actuality, it democratized the medium and allowed us to photograph everything. You can't shoot the Vietnam War with a view camera...you can't report on the flattening of the Middle East with platinum prints.

When I went to Colombia last summer, I didn't photograph much. As reported a few posts ago, I spent much of the time in the Library researching. I interviewed elder scholars from the area who shared meaningful insights--and shaped four narratives. I passed one of these to a friend from the area for an opinion, and he said he couldn't read it...since what I'd written hit too close to home. I'd had a similar feeling when I read Marquéz's Love in the Time of Cholera (which takes place in Mompox, a town near where my project is set), since the characters' lives closely mirror my relatives'. One of the characters in the book even has the same surname. 

I want to return to Colombia this summer, but the current political climate makes that unlikely. A colleague recently planned to visit Mompox, but was strongly advised against it. Recent instability in the area has made travel there unsafe, underscoring the distance between intention and reality. The impending presidential election there forces me to find ways to add to my project.

https://youtu.be/d45sda7WBs0

Enter AI. From the narratives I wrote, I translated them into Spanish. Ugh. Very stuffy and slow. I put them through a filter of Cordobés Spanish: the dialect of the area, so fast and so modified with colloquialisms that many Colombians themselves can't understand what's being said. I took out my digital recorder and mic and taped it. After I slapped together some footage, I ran my voice through an AI filter approximating a "northern Colombian accent". Not bad, but nowhere near as colorful as the speaker Martina La Peligrosa, right. 

But then, I replayed a segment from a spoken word artist (comedian, I guess...but more on him another time) and realized there's still a lot of work to be done.