AI: Its Problems and Potential

 

Image shows a black typewriter with a sheet of paper reading "AI Ethics" on a wooden desk

In recent weeks on Instagram, I've seen "whatabout-ism" around AI - "You're actually just racist, anti-Black, and operating from a colonialist mindset if you don't support AI",   "Neurodivergent people like me need AI to achieve our potential!"

LinkedIn is overwhelmingly in favour of AI, because of course it is - for people who talk such big games about how "business-minded" they are, how "hard" they work, and how "lazy and undeserving" people who are unemployed or under-employed are, the folks on LinkedIn, especially the white cis men, don't like to actually do work. They actually have a lot in common with the AI-as-a-disability-inclusion-right ADHD folk on Instagram; "I'm an ideas person; I shouldn't be hamstrung by the fact that there just isn't the time for me to get all my ideas out in a way other people can appreciate them!  Ideas are my currency, my potential - how dare people suggest I shouldn't use every tool available to me to achieve my full potential?!"

Let's break that down, for both groups - neurodivergent folk, and white, neurotypical cis men.

. I'm an ideas person - Not if you can't, through your own efforts, communicate and share your ideas you're not.
Anyone can have loads of ideas - putting those ideas into an actionable plan, that is comprehensible to other people, is the challenge - and the distinction between "people who have ideas" and "ideas people."

Why not co-create with humans? I literally have senior level experience writing both business plans and business cases. I'm an experienced freelance writer who has ghostwritten two novels, as well as having written and published my own novels, non-fiction, and articles; I'd be more than happy to take peoples' ideas, and, for a fee, work those ideas into something actionable. 

"I'm an ideas person, I need AI to realise my potential" is committing the following disservices to human beings:
. Keeping us from human connection

. Reinforcing hyper-individualism (which results in higher health costs, higher rates of absenteeism for businesses, and breakdown of social cohesion and social "agreeableness" - the ability to behave oneself in public, to know how to show up, when to shut up, and how to engage with people across all social strata who are unfamiliar to you.)
. Contributing to mental illness, physical health decline, stress and burn out by keeping us believing that our "potential" is only ever manifested through "productivity."
. Causing loss of self-worth, and the resultant costs in mental health support, social intervention, and lost opportunities, particularly for disabled and otherwise marginalised people - if we still can't "turn things around", even with these "fantastic tools", do we even deserve to exist in society?

Your potential is what you are capable of doing. Individually. Without tools which cause harm to environments and populations elsewhere (which AI very much does.)

A bedbound, non-verbal person, who isn't even able to log on to the internet, has potential - their potential may be creating paintings, greetings cards, or textile work. It may be providing a sounding board for others. Their potential can be found in the relationships that are present in their lives - both despite their condition, and because of it.

When we need to achieve goals which exceed our individual potential, that's where we should be working in collaboration with other human beings - which not only enhances the available potential, but also reveals potential which can be shared in other situations, builds social capital, reduces isolation and loneliness, and gets people out of their individual thought bubbles.

For people with ADHD, an essential life skill is learning to focus on one thing until that thing is done. Not "perfect", but present. As something that individual created, with their own competence and skill.

The claim made on Instagram was "But it would still be shaped by me" - but would it, if the issue is your ADHD "doesn't let me focus enough to get my ideas out there?"  Why would you suddenly be able to focus to "shape" AI's output?

Finally: writing, art, music aren't the sole ways to "get all your ideas produced" - a lot of ADHD people make really good YouTube hosts and convention speakers, because they can talk a lot, and engage people with a lot of tangenital ideas.  I'm really envious of that ability - I am verbal, but I can't communicate verbally in anything like an engaging way. I have presented at conferences, I've given sermons in churches (by invitation - I'm not a pastor), and I've presented to project teams and Trustee boards, but it's really draining for me. It causes me a significant amount of stress and anxiety.   Are those roles difficult to get, and highly competed for? Absolutely - but so are most roles these days. The UK currently has over a million more unemployed people than it has available jobs for them to go to.  And the more you insist that AI isn't a problem, the more of a problem with roles being competitive and difficult to get you create for yourself - because there will be fewer roles for humans.

. AI means marginalised people can compete on a level playing field - how? Seriously, how does "I can produce low-quality work in high volume, through a medium which is consistently devaluing human labour, and which you, my client can also access and use, whilst causing extensive environmental damage, as well as cultural harm, in exploited places, and lessening peoples' ability to think critically, and work collaboratively, which will result in higher health and social care costs on a global scale" give you any value to people you perceive as "only holding their value because of privilege?"    Yes, there are a lot of jobs I'm shut out of because I'm legally blind, a few more that are less accessible to me than they might otherwise be because of either my mental health issues, chronic illnesses, my inability to drive (courtesy of being blind), and/or the fact that my wife has supportive care needs (as distinct from physical/personal care needs) - by definition, the people who can take/do those jobs have the privilege of being able bodied, mentally well, and not holding kinship care responsibilities - that's their life. I don't resent them for that; they probably have their own challenges that I'm unaware of.  It is on governments, and organisations working collaboratively and systemically, to create a situation where there are equitable opportunities for people who do face systemic barriers. 

AI is causing significant harm to people even more marginalised than Western people who are non-white, disabled, neurodivergent, etc.

Most people are familiar with the water demand from AI's data centres (for comparison reference, "sustainable" Google or similar searching is <10 searches per day, "sustainable" AI use is 5 simple queries or 3 complex demands - ie, image/video creation - per day.  Because yes; Googling things uses a lot of water, too - hence why it would be really nice if people worked more on learning things for themselves, on getting comfortable asking other people, on learning how to research from books where they are physical able to do that, so they have less need to Google things.  I do not use Generative AI, and my average Google count is 3 queries per day; there are entire weeks where I don't have to Google anything), but other issues are that AI companies - the humans who own the technology - are deliberately and knowingly coercing vulnerable minorities in exploited economies into dead-end, low-wage, zero-hours contracts training AI - ie, working to make themselves obsolete - by telling them they're offering "jobs in tech" - which is technically true, but not remotely what is generally assumed by these people, who are desperate for "Western tech money", because it would make a genuinely radical difference to them. It would not just change their lives; it would allow them to have a life.

There are also talks going on about the future possibilities of "housing the data centres in giant satellites" - which would impact on the light from the moon, disrupting crop and migration patterns, and crashing global food security.  Plus, obviously, contributing to the huge amount of space junk that already exists.

AI is limiting human abilities that are going to be vital in an increasingly unstable and rapidly changing world - critical thinking, co-operation, group working, conceptual analysis.  This is making people a lot more controllable, influencable, and manipulatable - none of which are good things.

There are relevant, necassary uses for AI.  Some examples include:

. X-Raying and translating the scrolls recovered from Herculaneum - this is something humans cannot do; the scrolls are too fragile to sustain handling, transcribing and translating them all would take multiple human lifetimes.  However, the AI-generated translations will be checked by teams of highly experienced, professional human Classicists and linguists.
. Medical uses, particularly providing the equivalent of a "second opinion" in understaffed rural communities with high levels of underserved needs - doctors know they are not infalliable, and, in deprived areas, where medicine comes with a high cost, human doctors who do not have human colleagues to consult with are often dangerously hesitant, because they don't want to land a family with expensive treatments they can't easily afford.  Because medicine in non-Western, rural localities is not only expensive, but also can take days to get to, there is a far more pressing need than might be present in Western communities to catch life-threatening conditions at as early a stage as possible; medical-use AI can call on things like cancer and cataracts far sooner than a human could; in the West, doctors would hedge through repeated consultations over a period of time, so they could establish progression, and therefore, certainty; in remote, impoverished communities, where medical treatment is very much a luxury, that's not viable.

I am more than happy to use my own brain, produce quality work at a pace I can sustain, and find things out, and create, in analogue or analogue-ish ways, so that it is more sustainable for AI to be used to provide medical backup in vulnerable and underserved communities, and can give us the insights of vanished civilisations, as well as all the other not-fully-achievable-by-humans uses that AI can be applied to.

People will argue that "But they're different things! We're talking about LLMs! That's not the same as med-tech!" - but the resource demand exists for both.  The environmental and human cost exists for both.  Something has to be sacrificed - and it's far more realistic, and far more honourable, to acknowledge that people who don't want to/can't write books themselves, can't/don't want to create art or music themselves, can't/don't want to write a business plan, a CV, a cover letter, and who won't pay other humans, even in non-financial kind, to do those things for them, have no need to be doing those things than to insist that underserved communities go without prompt medical intervention so that you can claim you wrote and published a novel, or you can see yourself as a superhero. (I literally worked with a guy who can show you how to draw yourself as a superhero very recently -Stu Paterson. There are probably multiple human artists who can help you with learning how to draw yourself as a superhero. If you applied a bit of thought, you'd figure out how to do that yourself, potentially.)

"But it gets things done so much more quickly!" - and? You want to be paid less, because it takes you less time to do your work?  I can analyse a 52 page document and write a comprehensive report on it in 90minutes, covering multiple perspectives from competing stakeholders.  I don't feel the outcome would be enhanced by AI being able to do it in 90 seconds.  And I can answer questions directly on the report I wrote.

The fixation on "doing things more quickly" is contributing significantly to human health declines, to rises in stress-triggered autoimmune conditions, to mental health conditions. That is going to result in an unsustainable cost, in terms of health and social care, to countries around the world - and those governments are going to pass those costs on to their populations, in one way or another; populations who will be increasingly unable to afford those extra costs, because their job opportunities have been reduced by overuse of AI, and because they have compromised their intangible skills, such as critical and analytic thinking, co-operative working, and understanding of social dynamics, by their own over-reliance on AI.

"But I can't afford to pay for the stuff I use AI for!"
But you're paying ChatGPT?  If I could get 200 people paying me £10 a month to write anything they need writing, to work up their ideas, to suggest nutrition plans to them (I'm a qualified nutritionist, but I can't afford insurance to practice at the moment), I'd be happy, that's half what ChatGPT's typical subscription costs, I don't have a negative impact on exploited places and peoples...

You do not need to pay anyone - ChatGPT or a human coach - to "create a nutrtition and exercise plan" for you.  If you are able-bodied, that looks like:
- Move your body in ways you find pleasurable, for at least 10mins at a time, morning, lunchtime, and evening.
- Drink water at least once per day
- Eat fresh fruit and veg, as well as forms of wholefood protein you enjoy.
- Streching, resistance work, and cardio, with a little lifting, is most effective for average folks who just want their bodies to keep working.

A simple Google will tell you what hiring execs in your sector and area are looking for in a CV. That same Google search will often also give you templates you can copy and paste.

There are free image sites where photographers have been paid for their photos. You probably have a phone with a good camera.  You do not, in general, need either AI art or the skills of a graphic designer for most situations.

I have ethical objections to the common use of LLMs/GenAI on the grounds of the harm they do to exploited peoples and places, primarily, as well as the harm being done in Western countries by deepfakes and similar.

Ninety percent of the time, it's not the case that "AI stole my work!" - your publisher, or the social media site you showcased your work on, had a contract clause that said they were going to use that work to train GenAI. You just didn't read it. (As one writer to another - the publisher owns the rights to the book, not you. And yes, Amazon KDP is a publisher. You signed away those rights when you either accepted the advance or agreed to royalties.  Also, heads up, Vinted's new terms allows them to use everything you upload to train GenAI - I closed down my Vinted account as soon as I read that.)  However, it does concern me that AI is devaluing human creativity, at a time when creative professionals are already devalued.

I do not trust global governments, including my own, to support people who will find themselves permanently displaced from employment by AI. Governments and businesses are never going to tell people they can just live for free, governments are very resistant to the idea of Universal Basic Income. In the UK, as in many countries, common land which could be sustainably foraged has been lost. You would not be able to survive without some form of income - which would usually come from a job. But those jobs have either been lost, or devalued below a point of living income, by AI. If people aren't earning strong incomes which exceed the cost of living, governments don't make money from tax revenues. When governments don't make money, nation states fail, and their people are reduced to subsistence-level poverty.

My personal ethics and morality don't allow me to accept that level of risk of serious, life-altering harm for the sake of something I either don't need in my life, that I could acquire from other humans, or that I can do myself.

If you want to replace AI for working up ideas and writing, 

email me: theproductivepessimist@yahoo.com