The text began, “Hi Matt, I’m Pressella, a PR consultant at Pathos Communications.” It continued to tell me that an area nonprofit is closing down and in citing one of my previous job tiles said, “you’re a credible front-line voice on donor strategy and capacity , yadda yadda… We work pay on results, open to a quick chat?” It had my previous job tile and place of employment correct as well as my cell phone number – creepy. I didn’t click, I won’t respond, but I was curious.
Pathos Communications is an “award winning PR agency.” From their website, they state:
Pathos collaborates with its clients, comprised of SMEs and micro-SMEs, to create and distribute articles across a variety of platforms including established news outlets, digital media and podcast channels. This is supported by Pathos’ proprietary AI-driven technologies, PathosMind and Pressella, which are used to generate ideas, undertake market research and create news articles with limited human input required to generate highly efficient outputs.
limited human input, highly efficient outputs – sounds great. I’ll bring the matches, you bring the gasoline. I will absolutely not be helping Pressella as one of their limited human inputs. These companies and their optimization speak can’t even talk like a D-level HR flunkie with the worst people skills in the world. Have I mentioned that I hate AI?
A recently published report from the UN says that “AI-related water consumption could equal the basic annual domestic needs of 1.3 billion people by the end of the decade,” and could consume “945 terawatt-hours of electricity annually by 2030 – nearly triple the combined annual electricity use of Pakistan, Bangladesh, and Nigeria.” In this time of a worsening climate crisis, we need to be conserving energy, not consuming more. But hey, destroying the planet is a small price to pay for that AI generated slop video of the president flying in a fighter jet and dropping shit on protesters. Another report specific to housing in the US has suggested that the housing shortage in the US is being exacerbated by the rush to build data centers. Investment money as well as land that would have gone to housing projects is being diverted to the AI data centers gold rush.
There are lots and lots and lots of reasons to hate AI – from the environmental impact to the fact that it’s causing psychosis among some of its users, has told people how to carry out mass shooting, is responsible for generating a vast amount of misinformation, and has encouraged people to take their own lives. I hate it for all of those reasons…. but also because it is so emblematic of big tech’s move fast and break things mentality coupled with capitalism’s anti-regulatory, laissez-faire posture.
This, all of it, seemed entirely predictable and preventable. For people as “smart” as our tech overlords are, I’m frequently left wondering, why guardrails weren’t being put in place as the technology was being developed. A simple solution to the misinformation problem would have been to ensure that all AI-generated content (writing, music, or images) be labeled as such – much like we label genetically modified organisms. Moreover, to protect the rights of the artists, writers, and musicians (the “content creators”), there should have been an opt-out “code.” Meaning if I don’t want my content to be hoovered up in AI training models, I should have been able to put a piece of code on my site that essentially told the AI bots to fuck off and go somewhere else. (Actually, a best practice, a practice that respects human choice, would have been to set opt-out a the default – meaning you can use my content only if you ask in clear language not buried in some 19-page user agreement, and I say yes). Sam Altman of OpenAI (Chat GPT) said it would have been impossible to train AI without using copyrighted material. He also said a “revenue model that compensates artists for the use of copyrighted material would be ‘cool.'” Thanks Sam, that’s generous of you. This admission is, of course, after they’ve already used such material. As critics of AI have suggested, if your business model relies on stealing the work of other people, perhaps you need a different business model. But that is not how “tech” typically works. As we’ve seen time and time again, they operate with an ask for forgiveness later (if we’re caught) mentality.
I have many days when I ask myself, why do I even care? Put simply, it’s because I also have many days when I’m certain that my opposition to AI will impact my ability to earn a living or retire. And even if I could retire, I worry that AI is destroying our ability to recognize and appreciate human-generated art. A writer I follow recently conducted an experiment in which she submitted one of her essays to an AI detection site. It said her writing was “84% AI written” and the offered her the opportunity to pay to have it “humanized.” So if I understand this correctly, they’re basically saying, we’ve poisoned the content environment, but for a small fee, you can have the antidote. I really hate that so many people in the tech industry seem to have little regard for consent and that as someone who doesn’t consent, I have few options available to me.
And don’t get me wrong, I have no delusions about the value of my writing or art, but even so, I don’t like the idea that AI companies are scouring my blog, my social media, and my LinkedIn profile to either train their models or better scam me and sell me shit. In the last two years alone, this site has seen “page views” jump from about 20 or so a day to between 60 and 90 views per day (this week’s average is over 155/day) with several spikes in excess of 500 page views. I have no doubt that my resume and cover letter along with my poetry and other writings are being (or will be) flagged as AI, and as someone who avoids using AI, the whole situation is disheartening, if not infuriating.
I would love to end this post with a solutions-focused, John Oliver style, “so what can we do?” For me, I probably need to read more optimistic articles suggesting that none of this is a foregone conclusion. I should probably look for those articles that say the bubble may burst, but it won’t take the entire economy down with it. For the time being, I’m pretty much relegated to riding it out. Try to earn and save enough money to insulate myself against it, or go into a recession- and AI-proof business like bar tending. But even those options feel like a long shots that’s only growing longer. Shrug.