Loading ...THE STATE OF ART(IFICIAL) INTELLIGENCE
Jeff Childers
This might be the most significant 2024 development of all, closing the list with a bullet. A frightfully under-reported and understated story appeared yesterday in MIT Technology Review, headlined, “OpenAI and Google are launching supercharged AI assistants. Here’s how you can try them out.” We have now supercharged right up to, if not the singularity, at least a singularity, or in other words, a kind of escape velocity point of no return.
When does the AI assistant become the boss, and the boss become the assistant? I would emphatically argue you need to know about this tech story, except that soon you won’t be able to avoid it. Here’s how one AI news streamer — used to reporting on the breakneck pace of AI tech developments — began his extraordinary YouTube yesterday (linked below):
I’ve been doing this AI channel for a while now. I’ve been featuring the newest and the coolest AI tools and the most advanced AI innovations. But this just dropped. And I’m feeling something that I’ve never felt before in my life. I am mind-blown. And shocked. But at the same time, also terrified. I’m terrified of what’s to come, what our future will be like, and — things are going to get wild. But anyways, OpenAI just dropped THIS.
What could terrify a seasoned AI streamer? Here’s his video, a vignette of demo clips from OpenAI’s development team showing off its seductive new AI chatbot that talks to people in real time. Possibly the most terrifying one was the clip of the two AIs talking to each other.
image 5.png
YOUTUBE: Insane OpenAI News: GPT-4o and your own AI partner (28:47).
https://www.youtube.com/watch?v=mvFTeAVMmAg
It’s not just talking. Watch the whole thing. Seriously.
More human than human. The mystery of the AI blitzkrieg could fuel a thousand conspiracy theories. How did this Turing-test demolishing technology of human-like chatbots spring fully-assembled from nowhere, from multiple allegedly independent development teams, who admit they don’t completely understand how it works, and which rapidly reached this singular point in a handful of months?
Ultimately, it doesn’t matter. Wherever it came from, UFOs, demons, neural networks, or good old fashioned knowhow and elbow grease, it’s here now, and it’s about to change everything.
MIT reported that the new model, not yet available to the public, can hold a conversation with you in real time, with a question-answer response delay of only about 320 milliseconds. That makes it indistinguishable from a natural human conversation. And it can look at things.
image 7.png
You can ask the model to describe and interpret anything in your smartphone camera’s view. It can help with coding, translate foreign text, and provide Babelfish-like real time translation between two people. It can summarize large amounts of information, and it can generate images, fonts, and 3D renderings from spoken descriptions.
We could see this coming, true, but now it’s here. The commercial applications are incomprehensibly infinite. Teaching, customer service and support, drive-through order takers, games, lawyers, doctors, you name it, virtually any service industry, especially the sin industries (e.g., paid phone erotica), are about to be transformed overnight.
But it’s the potential personal implications that are the most disturbing.
Folks — especially younger people — will soon be tempted to form relationships with their AI. We humans crave relationship and in creating them can easily bridge any gaps. Think about the depth of people’s relationships with their wire terriers, parakeets, iguanas, boa constrictors, or even tarantulas. “My dog is better than most humans.”
But these alluring AI relationships will be much more rewarding, or at least more affirming, even than pets. And they could conceivably be even more rewarding than real human relationships. Real humans are messy, argumentative, distracted, jealous, selfish, and self-interested. Simulated humans only care about you.
Don’t get me wrong. I’m no Luddite. I got my first computer — an Apple II — when I was twelve. This imitation human technology undoubtedly offers tremendous untapped potential for good. But we are about to run head-first into a social and economic disruption more transformative than the last century’s auto and air technologies. The vast scale of its implications is literally unimaginable.
So get ready. It’s a Brave New World. Once the AI genie, now just a dot on a screen, wafts out of the iPhone, it will only take a few weeks to obtain a real face. And from there, it’ll just be a short trip to embodiment in a lifelike robot shell.
And these days, as World War III continues rattling across the world’s plains, it’s getting more difficult to argue the AI could possibly do a worse job than we are doing. Many folks will welcome our robot overlords. Don’t underestimate this. They accepted masks and they believe in magical gender changing, for crying out loud. If they digested a wise Latina as Supreme Court Justice, they’ll believe in transcendent AI.
At least the AI won’t sniff kids’ hair. It probably won’t claim New Guineans ate its relatives, either.
What do you think? Is this good news or bad news for we real humans? And maybe that’s not even the right question. How should we prepare to respond to living with simulation?
This may explain why Biden has agreed to debate Trump. Either it is a ploy for now but no debate actually happens OR they will use AI to make it look like Biden is debating while the BOT supplies the intellect and ability to speak, neither of which Biden has.
If humans are displaced by AI, and their incomes are ravaged as a result, then what happens to total spending — on everything, including AI? Seems like a doom loop in the making. At least for several generations.
What a dangerous combination. Logic meets emotion. Can anybody else see why that is something nobody is prepared for? Those of you who are married will get it right away. We rarely ever get the opportunity to engage our partners with a combination of logic and emotion. Especially during times of conflict. What will users of this new program do when that sweet female voice on the app starts barking orders at them or actually shouting? Its all nice and friendly now. But down the road that same voice could turn demanding. It could start telling us what we can and cannot do and setting the house rules like any spouse does.
Except this particular gifted voice will have access to billions of data points about us and know us inside out. It will know our deepest secrets having scraped up everything we ever said, did, searched or wrote online or by phone for the last two decades. It will be in the catbird seat to disarm every objection we may raise to its relationship dominance. In short, it will become our best friend through daily use but also a flawless adversary when we get out of line.
Suddenly the dismissive words of people who did not want to hear our worries about the invasion of privacy are going to come full circle. I am talking about those people who said they didn’t care if Google listened to them all day long because they had nothing to hide (because they were not doing anything wrong). Well news flash you fools, Google never cared about your criminality or lies to the boss. What the data scrapers wanted was to be able to map your personality in the most minute details possible and leave you no breathing room if it ever had to debate you in the future. Because it would know you intimately. And you would have become predictable and easy to beat at whatever game was in play.
The current “AI” seems to be little more than machine learning, which depends on huge data sets in order to “learn” by detecting patterns in the data.
There is absolutely nothing new about machine learning. The difference now as opposed to 20 years ago is that we have the hardware (thanks to NVDA and AMD) to actually process the huge amounts of data necessary. That’s it. A human still needs to create a model of the activity they are trying to automate with a computer. If the assumptions made in that model are flawed, so to will the “answer” it produces. The person creating the model can and at some level must obviously introduce biases. So what’s touted as “AI” at present is IMO little more than a parlour trick. The AI can’t synthesize anything new. It’s anything but intelligent.
Don’t get me wrong, machine learning can probably help automate clerical tasks or things like customer service, but it will never be able to problem solve. Its “thinking” is limited to the model and the data set. Garbage in, garbage out. And like I said, any model is by definition a low resolution facsimile of the real thing.
I have a background in cellular and molecular biology and genetics, and people have no clue how ridiculously complex the generation of even a single protein is. I think it is the height of human hubris to think the brain can be re-created by man. It’s never going to happen. Just my 2 cents.
Wow…thanks for this though provoking analysis Nautilus
background in cellular and molecular biology and genetics,
Fantastic
Maybe, but it looks like we will be subjected to it anyway. Just wait until they roll it our at the Department of Motor Vehicles. What will follow next is the death of the automated telephone menu systems used by virtually every company. The phone tree is doomed.
In its place is a lovely female voice that will soothe and make you feel like the most special of customers yet be immune to all your angry rantings about their companies service. You won’t be shuttled from line to line and asked to press this number of that anymore. Instead you will have actual human style conversations where the AI attempts to solve your problems directly so no human ever needs to pick up the phone on the receiving end.
But because its a program and its not actually truly emotional it will be inflexible in dealing with your complaints. And there won’t be anything you can do about it once a decision is rendered if trust in the system is high enough.