Top A.I. Engineers Warn Humanity
“Potentially resulting in human extinction…”
OpenAI Engineer William Saunders and 17 other engineers of OpenAI and Google DeepMind recently penned an open letter to Humanity warning us all that this generation of A.I. tech is woefully misunderstood and very dangerous. ( Full Letter Here.)
Let me illustrate what is going on here, in very clear terms so we can grapple with this together.
To start at the beginning…
I’m excited about this generation of A.I. and the enormous evolutions it will bring to our world. And when I say enormous, I mean it…
And in a short amount of time, I mean 5-10 years. The evolution we will see in the next 5-10 years is equivalent to the evolution we have recently seen in the last 50 years. (Pause for a moment so you can take that in. 5yrs > the last 50 years!)
So that includes our evolution via the internet, smart phones, personal computers, affordable home solar systems, online streaming, the end of Oprah’s daytime TV Show (you get my drift!). The advancements in the last 50 years have changed our global language, culture and society. From one perspective this changed our world from a society that valued ‘information’, like encyclopedias, libraries and schools as ways to discover facts and information, into a society that values ‘attention’. Because now all information is available and it’s what is going to garner your attention that has value.
Before A.I: Society values =’s facts and information
After A.I: Society values =’s attention
But knowing how to curate, how to communicate, and how to evaluate information or attention, is still very much underutilized by our society as a whole. Valuing attention over information has come on very fast and because we don’t know how to curate or evaluate, we are still coming to terms with the full impact of this change on our culture.
For instance, in studies on social media we have seen elections won, and human violations and regimes that support those violations toppled via social media. The poorest of our population gaining access to smartphones and then getting medical and financial help via crowdfunding sites. And a global pandemic peacefully navigated (in Taiwan and New Zealand) and less peaceful (USA and Europe) due to digital technology and social media. But also, the advancement of gender fluidity and LBGT rights. And then those rights taken away, discriminations, elections saturated with misinformation and bias, and the largest suicide rates in children under 15 ever recorded, for example.
The last 50 years created an enormous amount of access… Access to information, abundance, culture, and diversity. We have a new kind of celebrity now called ‘An Influencer”, we have a different education, see children don’t learn cursive writing anymore, or long form math’s. We are different people, see Millennials (first gen of attention society) vs Gen X (last gen of “information” society). But 50 years isn’t long, so we are still learning to see and evaluate the impacts. But stand with a group of 17 year old’s talking to each other and see if you can understand the language or nuances, see if you don’t feel like a visiting alien to that species.
But this is the kind of evolution we will see over the next 5-10 years.
Because the access will be to the power of 100. Because this generation of A.I. is about creation, not curation. And this is very hard to imagine. Very challenging to vision the course these changes will make to society.
Google Search, Slack, Amazon, Reddit, Facebook, Ticktok, etc=’s Curate information
OpenAI, GoogleDeepMind, Microsoft’s CoPilot, etc =’s CREATE information
Right now, when you type something in to google it curates what you can go look for. Your friends and family can curate, or recommend things to you that you might like, and what your influencers curate (like me) etc.
But creation is different.
Remember the movie HER, 2013 Spike Jonze romantic drama? In the movie the computer operating system is very human-like,instead of Windows or Safari its a HER, a very sexy and alive computer learning system. The main character falls in love with HER and has a very interesting and very human relationship. This is what creation is closer to.
Within 5 years we will all have HERs in our pockets. A friend who can create anything for us. Like a course on how to overhaul an engine. NOT curated articles, but a very alive, maybe even sexy, coach who is right in your hand, walking you through how you personally can learn to do this. Or how to install your new washing machine, or lay bricks, or follow me here, a Non-Dual coach.
You can have a HER who can take on the shape of a non-dual coach and you can say you want to have her coach you in the style of Eckart Tolle, or the Buddha. And over the next 3 days, you want your HER to walk you through a retreat based on these teachings. Additionally, you want your HER to correct you when you are suffering or causing pain and help you become enlightened.
Cool, right?!
But then you might decide that you “got it!” maybe a few months of these retreats, and you have a better solution than Eckhart. So, you ask HER to create a podcast, say 30 episodes on your insights and in the podcast include celebrities and influencers who agree with you, and then also a documentary about your insight, including in the documentary Eckhart himself agreeing with you and let’s include Rick Archer hosting the documentary. And also, a course about your insight that you can sell online. And guess what…HER could create those things for you in about 2 weeks. (Without any of the celebrities or Rick Archer, etc ever knowing.) You can almost have that today, but in 5 years, you most certainly can have that without your audience of these offerings being able to know if any of it is real.
Right now, A.I. can create multiple celebrities wishing you a happy birthday with private messages. AND In their voices, with their images and in their own homes! And you cannot tell what is real, not even those celebrities can tell.
Keanu Reeves Deep Fake
Now I do not fear that I will be out of a job, in fact I think the opposite, real human wisdom will become very, very valuable. I suspect the created materials will only serve beginner levels, because deeper wisdom requires a personal nuance, so this will become hugely important in the next 5-10 years, with much larger numbers of people needing more nuanced and individual support. However, perhaps you can start to sense what those engineers are saying by “potentially resulting in human extinction”. I mean I am talking about a Non-Duality farce, which, let’s face it, we already have tons of this going on. But what happens when it’s about weapons, war, genocide? What if a 25-year-old man who didn’t have the best upbringing gets a gripe with Asians and gets his HER to ……fill in your own blanks. Because right now, we have a few dozen 20–30-year-old white guys in Silicon Valley who have influenced the entire world by creating an algorithm for Facebook and Google that has zero consciousness or consequence or pause and has led to some disastrous results on a global level. No offence to 20 something white guys, but they were not appointed to run the world, and the mass suicide rates on our young people prove they were not qualified. And I don’t want that 25-year-old with a gripe doing it next.
So,
Although I am an A.I. fangirl, I think it does need to be paused and not raced forward. Because we don’t yet understand how good it can be, and how limited it can be, and when we are blind to limitations, we run into problems and harms we could be free of.
Currently we have a real challenge pointing out misinformation, and where it is veering off into inaccurate and/or harmful. This is really true with “awakeness.”. A leader can be 80% accurate and 20% inaccurate, but that 20% is really subtle, but can make the whole of the teaching inaccurate. And we as a spiritual or Non-Dual community have not even created language models ourselves for this. Meaning we haven’t found collective words to define awakening, or language for misinformation around awakening. So, we can’t share language models with A.I. creator bots and expect accuracy of our materials in the A.I. databases.
For instance, I have a lot of free materials, courses, blogs, YouTubes, etc. These materials can be used by any A.I. bots saying the “right” thing, however subtly confusing what is real and then leading people to trance states, not source states. Which is a very hard thing to language and to point out. Awareness is also a very subtle sense, and we often allow inbuilt bias, more accurately called discrimination, to exist without us really seeing it. In the “awake” leaders space we have very few female leaders, and even fewer people of color .
And the way that A.I. data is gathered, they call it mined, is from open sources. Meaning all those free materials are being scanned, this is “machine learning”. A.I. bots gather all information online including social media sites like Facebook, YouTube, Instagram, public websites and blogs, etc. And in every single one of these gathered sites there will be a bias, and it will be inbred in our language, in our representations and in our online images. This creates bias in the A.I. data that has a broader impact when this tech is not ‘curated” information for us, but “creating” information, creating relationships, creating entertainment, creating learning, etc.
The 2 biggest challenges with this generation of A.I. is:
- A.I. can’t learn what we have not learned yet: how to curate, communicate, and evaluate. And this generation of A.I. will impact us so greatly we may become blind to ever learning these ESSENTIAL tools.
- A.I. will not eradicate bias, but instead deeply amplify bias. And bias is subtle, so even when amplified it may not become obvious, but just more broadly adapted.
But the generation of A.i. does not need to be released at this speed. It can be paused until it is better understood. We can absolutely put into place all kinds of checks and balances, and we can bring a lot more global awareness towards the new technology wins and losses BEFORE it’s globally available. Last summer the movie Oppenheimer, 2023 Chris Nolan drama, was released, rumor has it that it was created for the A.I. engineers to learn from Oppenheimer, creator of the nuclear bomb, and the pressures, and regrets.
PAUSE. PAUSE. PAUSE.
We can pause.
The only reason NOT to is financially, whoever gets to market first wins the $$. And political competition, beat the Chinese Government to the creation tools so it doesn’t overwhelm the political power base, by that government having so much creation tools it wins the world domination race.
Both of these are compelling reasons to blaze ahead. But are they more compelling than “possibly resulting in human extinction?”
No.
So, please don’t jump on the bandwagon too fast with the clever A.I. options from google, or Amazon, or for your business, etc.
Be aware, be clear, and be slow towards the use of this powerful,yet hugely underevaluated, tool.
XOK