Generation AI Part 1- Why We Should be More Afraid of Techno-Oligarchs than a Rogue AI (for now)
Generation AI
The Greatest Generation had the Great Depression and World War 2. The Baby Boomers had Woodstock. Generation X had the Internet. The Millennials had social media and smartphones. Every generation has its defining moment. For those born within the last few years, it will undoubtedly be the arrival of artificial intelligence. So, maybe it should be dubbed Generation AI?
In the same way that most millennials do not know a world without the Internet, Gen AI will never know a world without artificial intelligence. Pundits are saying that its arrival is as significant as the invention of the wheel, the discovery of fire, and the creation of atomic energy.
Nevertheless, Geoffrey Hinton, widely regarded as the godfather of artificial intelligence (AI) quit his job at Google, warning about the growing dangers of the burgeoning technology. In an interview with the BBC Hinton said that AI chatbots were “quite scary” and that “Right now, they’re not more intelligent than us, as far as I can tell. But I think they soon may be.”
While I agree that AI could eventually pose an existential threat to humanity, for now, it requires prompts to function. Yes, at some point, if artificial intelligence becomes autonomous it could wipe us out, but right now, as we’ll discuss below, it is still incredibly reliant on humans. So, for the time being, I think there is less to fear from an AI gone rogue, and more to fear from the humans who are directing it.
Let me begin by saying, I am not a Luddite. I love technology. I built my first website back in 1997 and way back then the media said that the Internet would be a great equalizer that would create a level playing field for small companies to compete with large corporations. They said it would free us from the shackles of our desks and allow us to work less.
While some of these things have come to fruition, many have gone in just the opposite direction. Today digital nomads roam the world while working from exotic beaches in Thailand or Bali, so yes it has unshackled us from our desks. But we are working more hours than ever. Thanks to email and text messaging our bosses can get us any time of day or night even on the weekends.
The Internet has made it easier for entrepreneurs in niche markets to start businesses, but the rise of tech giants has meant that small and medium-sized companies have been crushed or bought out creating consolidation across many industries.
Far from being the great equalizer the majority of the spoils have gone to a handful of companies, in particular, Google, Microsoft, Facebook, Amazon, and Apple which have to some degree hollowed out the middle class through AI automation and have even wiped out a few large corporations along the way(Toys R Us, Bed Bath and Beyond, Macy’s Nordstrom, to name a few).
When it was just blue-collar factory workers, we weren’t so alarmed but with the release of Chat GPT even white-collar workers with MBAs and PhDs are starting to get nervous.
The Internet has destroyed the small local newspaper that depended largely on classifieds for ad revenue. Since 2005 2,200 have closed. Much of the money that used to go toward ads in newspapers and magazines, and commercials on television has gone to search and social media giants Google and Facebook.
More recently, digital news media companies have broken down, including Vice, and Buzzfeed News, and Vox recently laid off 7% of its staff. The Internet is even threatening to topple large news organizations like the New York Times, NPR, CNN, and Gannet.
So, the question is what will generative AI mean for the beleaguered journalism industry? In 2020 Microsoft replaced the reporters that maintained its MSN web page with AI. CNET has used it for tech reporting, although it was plagued with errors, and Buzz Feed used it for its travel section.
But it’s not just print journalism that is under threat. AI can even replace news anchors and reporters with digital doppelgangers that are hard to distinguish from humans, in the near future it will be all but impossible.
Google announced in May 2023 that they were making plans to further undercut the news organizations with the release of Bard, their new AI-assisted search engine which will summarize the top search results, alleviating the reader of the need to go to the site. This will mean fewer eyeballs for organizations like the New York Times and less ad revenue.
For decades the New York Times has been seen as the pinnacle of serious journalism. If we lose these types of news organizations it will mean a huge loss for democratic societies, which depend on them to act as a watchdog for corporate and government malfeasance.
What will happen to investigative journalism if we put it in the hands of an AI? Can we trust it to hunt down leads and break historic news like the Water Gate Scandal, the Pentagon Papers, and Exxon lying about climate change?
Has Social Media Made Us Better Off?
AI is supposed to democratize knowledge and information, but how can it if so, few people control it? Just five (Apple, Google, Facebook, Microsoft, and Amazon) tech companies control most of the artificial intelligence market in the United States and the data that it needs to run. This concentrates, power, money, and influence in fewer hands.
Ask yourself if you feel like our society is better off than it was 10 to 15 years ago. By some metrics we are. Our computers and devices have gotten faster and cheaper, and so have telecommunications, but it has come at a cost.
In 2019 the U.S. had the highest suicide rate it’s had since World War II, increasing by 17% since 2000, even while it dropped by 36% in the rest of the world. In 2021 the U.S. had over 100,000 drug overdose deaths, the highest ever recorded ever and up 28.5% from just the year before, according to the CDC.
We are seeing higher levels of anxiety and depression; young adults are having less sex than 20 years ago; mass shootings and homelessness have risen sharply in the past few years; and the Surgeon General just declared a loneliness epidemic in America.
All these issues can at least partially be attributed to the rise of social media, the Internet, and artificial intelligence. These technologies were supposed to bring us together, and while it is easier to stay in touch with friends that you have not seen in decades, we are also more divided than ever. The spread of misinformation, cyberbullying, job loss, and the concentration of wealth has also led to an angrier country.
Although wealth inequality has fallen slightly since 2000, today a mere 1% of people control 46% of the wealth while the 70% at the bottom control a scant 2.7%. Things will get worse if we do not regulate AI and truly democratize it. Yes, AI can be an incredible source of good and could potentially end poverty, improve healthcare and education, and help end the climate crisis and biodiversity loss.
But just like social media, it can lead to a greater spread of misinformation, only this time it will be much more difficult to distinguish the difference between lies and reality since using AI perpetrators can manipulate us with deep faked images, videos, and scientific evidence making it a source of tremendous potential evil.
The rise of AI could mean the end of democracy as it would be so easy to manipulate the news and elections. It will most likely lead to even greater concentrations of wealth as billions are bilked out of their savings in complex scams and the entire world becomes dependent on AI for basic needs such as healthcare, education, and housing.
Technologists are talking about the potential of AI replacing artists, journalists, filmmakers, actors (yes actors), scriptwriters, authors, teachers, doctors, lawyers, judges, and even parents. We are trying to be sold on the fact that AI runs on algorithms and because of that they are unbiased, but that is not even remotely true.
AI is only as good as the data it is trained on and in real-world usage, it is every bit as racist, sexist, elitist, and even misogynistic as humans can be, perhaps even more so since it’s only such a small part of the population that controls this technology.
So, imagine a world in which, every person in one of these jobs is a racist, sexist, elitist, misogynist and you can get a picture of what we could be creating if we continue along the unregulated path that we are currently on.
Imagine if every movie you ever saw, book, or article you read; every time you stepped into a classroom; a courtroom; or a doctor’s office you knew that the person who you were interacting with had a bias against you. It would be like stepping back into a world of slavery, except this time most of the world would be the slaves and your masters would also know what you are thinking and feeling.
A young, attractive, influencer named Caryn, created an AI replica of herself to keep men from being lonely. You can chat with it for the sum of $1 per minute and it will even talk dirty to you if you are into that sort of thing.
Humans like computers are programmed. But instead of 1s and 0s we are programmed by our hormones, parents, teachers, friends, the media, the books we’ve read the movies and TV shows we’ve watched, the traumas we’ve experienced, and the kindnesses we have been shown. The result is that human beings are sometimes unpredictable and oftentimes irrational. What is considered normal behavior to one person is considered aberrant to another.
I’m an introvert by nature and know first-hand that humans can be a phenomenal pain in the ass and are often disappointing. But the difference is that we can disconnect from one person and move on to the next.
What if we become so dependent on AI companionship that we start losing touch with our human friends, or worse forget how to communicate with them? Do you think that is unrealistic? Again, young adults are having less sex, because online porn has become a replacement for human relationships for a large swathe of the population. How big a leap is it to go to an AI robotic companion that on some level seems to fulfill your every need because they are so good at manipulating you?
Yes, as humans we manipulate each other as well. But with one very big difference. An AI will be able to read your thoughts and know what you are feeling based on your blood pressure, heart rate, and breathing patterns before you even know it. We could eventually live in an age where we are so dependent on AIs to fulfill our every need that we have no choice but to pay big tech companies whatever they are asking for them and it will probably be a lot more than $1 per minute.
What happens when AI and the people that control it are responsible for every TV show and movie you ever saw, every book that you ever read, everything you were taught in school, and most friends you have had were made by an AI? How differently would you think, act and behave? Would you even have any thoughts of your own that were not indoctrinated into you?
AI Isn’t as Smart as We Think It Is (Yet)
In a TED Talk titled- Why AI Is Incredibly Smart and Shockingly Stupid, computer scientist Yejin Choi demonstrates how little common-sense Chat GPT 4 has when she asks it two questions that a six-year-old child could answer. The first is “If you have 5 pieces of clothes and they took 5 hours to dry in the sun how many hours would it take for 30 pieces to dry?” The answer Chat GPT gives is 30 hours.
Then she asks “I have a 12-liter jug and a 6-liter jug. I want to measure 6 liters. How do I do it?” The bot responds by spitting out a lot of complicated nonsense.
Right now, AI is a bit like having a child with a genius-level IQ, but no emotional intelligence. Yet it is already deciding who gets mortgages, and who gets hired and it has led to charges of discrimination.
Chat GPT is also not as independent of human intervention as you might think. What makes the responses of the inquiries that you input into the bot sound so human, is an army of $15/hour grunt workers labeling photos, videos, and chunks of text to help the AI recognize, learn and respond to them. Like a clothing brand that outsources its manufacturing to an overseas company, the workers have no benefits or healthcare, and the work is on demand. While the spoils and the glory go to the executives above.
AI Needs Regulation
Eric Schmidt the ex-CEO of Google in an interview with Meet the Press said that we should allow the technologists to regulate themselves and set their own “guardrails,” when it comes to AI because government bureaucrats wouldn’t understand it anyway.
Should we trust the heads of these technology companies, many of whom are libertarian and resent any oversight at all? Again, we need to ask ourselves if we are better off than we were 10 years ago. On most metrics, I would have to say no.
There needs to be an AI agency whose sole job is to monitor and police AI. They should not be a toothless agency like the SEC which can only hand out fines, but they should be more of a policing agency like the FBI which can give jail time.
The agency should be headed by computer programmers, engineers, technologists, as well as academics, philosophers even science fiction writers who can anticipate the worst nightmare scenarios and prevent them from occurring, and yes, they should even have their AI that is trained specifically on ferreting out crimes related to AI.
There is already a blueprint for an AI bill of rights, that puts restrictions on data use, discrimination, and safe usage, but this needs to be fleshed out more and turned into law. Even if they will be hard to enforce, it at least draws a line in the sand so that we can delineate what is on the right side of the law and what isn’t.
There need to be restrictions on what jobs AI can replace, or at the very least there should be oversight on them. Investigative journalism still acts as an important watchdog to curtail the abuse of power. This should not be left to an AI.
To this the technologists would answer, if we do this we will fall behind in the development of AI. I say, so what. China probably already has surpassed the US in the development of AI based on the number of patents they have filed, the number of papers they publish, and the fact that they have far more data on its citizens than any country in the world probably does.
Having control over massive amounts of data may make for a better AI, but does it make for a better world or a society that we want to live in? That is the question we need to keep asking ourselves as we develop this technology.
Google/Alphabet, Apple, Amazon, Microsoft, and Facebook have all become monopolies. Any business that has the reach that each of these companies has, into every single facet of our lives is by definition a monopoly. They need to be broken up and prevented from reconsolidating in the future.
AI Requires Greater Transparency
Right now, only big tech companies can train AI because of the enormous amount of computer power, data, and human labor that it requires. However, there needs to be greater transparency and oversight as to what data is being used and how it is being used. Perhaps a citizen lobby made up of normal people could create a more democratic way of deciding the direction that AI takes.
A Tax-Payer Funded Media
Good journalism is as essential to a fair and open democracy, as voting is. While AI might be able to write it is not a journalist who is concerned with reporting the truth. One way to ensure that journalism survives could be to tax the search and social media companies and reallocate this money toward rebuilding local newspapers and creating other independent media outlets that are not overly reliant on corporate ad revenue while remaining independent from government interference. This would be a truly independent press.
The U.S. was a country that was founded on the principle of living free from tyranny. In the almost 250 years since its founding, it has not always held to those values, it has nevertheless remained a beacon of hope for refugees fleeing oppressive societies. However, if we allow AI to take over all facets of our decision-making, instead of being a country ruled by wealthy landowners, we could very well become a tyrannical state ruled by wealthy techno-oligarchs enabled by a very powerful AI.
Originally published at https://palmerowyoung.me.
