Ray Kurzweil’s Prediction Scorecard Now That ChatGPT Is Here
The futurist famous for projecting the singularity by 2045 has a track record. With generative AI now making waves, the hits and misses of his predictions may hold some insight as to how things might change over the next few years.
20 years ago, how close would you have predicted the ways in which computers and technology are a part of today’s world?
Would you have foreseen the social media boom? The smartphone? The drastic increase in people working remotely?
Well about 20 years ago, a brilliant computer scientist by the name of Ray Kurzweil did make predictions on the future of tech. In fact, he made well over a hundred of them. He also made some bold predictions about what should happen in another 20 years.
Kurzweil’s foresight on the future of computers has received a lot of attention over the last couple of decades. Because he put all these prognostications out there, he sort of made a roadmap for us to assess our technological progress.
But he also gave us a lot of prediction data that might not just hold lessons on what track we’re on. They might give us an understanding of why some tech forecasts pan out and some don’t.
Today, Kurzweil is the Director of Engineering at Google and arguably the most renowned futurist of our time, especially when it comes to computer technology and artificial intelligence (AI). In his 2005 book, The Singularity Is Near, he famously predicted that by 2045 the technological singularity will occur.
Kurzweil defines the singularity as the moment when AI vastly surpasses all of human intelligence. And by that he meant Artificial Intelligence will think better than us, create better than us, and make better judgments than us in pretty much every aspect you can imagine.
And now, we are essentially in the middle of Kurzweil’s 40-year singularity timeline (2005–2045).
Maybe then it’s appropriate that at the moment the world is currently encountering ChatGPT. ChatGPT is a high-performance generative AI and large language model, which means it creates written content based on prompts. You give it questions and it gives you in-depth, thorough answers that feel almost like you’re chatting with a human. This certainly seems like a development an AI futurist would have anticipated.
Now, in looking back at Kurzweil’s predictions for the last 20 or so years, I think we will see some clues as to how the advent of generative AIs will influence our world in the near future.
CAVEATS AND STIPULATIONS ON THE “PREDICTION SCORECARD”
- Kurzweil made many predictions. The ones I selected were those I felt were most relevant to our relationship with AI and computers in general.
- While I did provide supporting links in many places, some predictions I felt were obvious. For instance, we all know we don’t have personal flying cars yet. We also know smartphones are something nearly everyone has.
- While Kurzweil made his “singularity by 2045” claim in 2005, the bulk of his projections came from The Age of Spiritual Machines in 1999. So unless I note otherwise, assume 1999 is the year when the prediction was proclaimed.
WHAT KURZWEIL GOT RIGHT
FROM “SPOT ON” TO “JUST BARELY”
PERSONAL COMPUTERS WOULD FIT IN OUR POCKETS
Kurzweil predicted that by 2009, PCs would be “available in a wide range of sizes and shapes.” And lo and behold, by 2007, we had the iPhone. By 2008, we had the Fitbit. If we look around the world today, small computers are abundant. I think it’s quite obvious that Kurzweil nailed this one.
THE CLOUD
In 1999, Kurzweil was completely accurate when he said that computer devices will one day link up and “become communicating Web servers, thereby forming vast supercomputers and memory banks.” He knew that this would allow for “high-speed communications and computational resources.” He called it “the worldwide mesh,” but we know it as “The Cloud,” and it is a dominant force in computing throughout the world.
TALKING TO A COMPUTER
Kurzweil basically foresaw that we could verbally ask computers simple questions and get somewhat reliable answers back. He also thought they would be able to tell jokes, schedule reminders, conduct transactions and more. While humans may be the ones coming up with the witty remarks for Siri, Alexa, and Google’s Assistant, the interactive qualities of these AI communication systems are very real. In fact, it’s so real, 4-year-old kids are surprising their parents by talking to smart devices and ordering fun toys for themselves.
MILITARY DRONES INCREASE IN USE
He thought that by 2010, the military would be relying heavily on unmanned airborne devices. That premonition has definitely come true as drone strike usage surged in both Obama’s and Trump’s presidencies.
A UNIVERSAL TRANSLATOR
The futurist thought that by 2009, “We’ll have real-time translation of foreign languages” and translating telephones. The tech has been around for a while (Google translate did come out in 2006), and it’s only improving. It may not be used by everyone and everywhere like Kurzweil guessed, but if you’re smartphone savvy, I’m sure you could find an app to communicate in any language.
GOODBYE CDS/DVDS — HELLO STREAMING/DIGITAL
Kurzweil anticipated that everything from songs to movies would usually be acquired digitally. Today kids don’t know what CDs are, video games are downloaded, and blockbuster movies are streamed often on the same day as their theatrical release. Case in point — DVD sales dropped 86% from 2006 to 2019.
PINHEAD-SIZED CAMERAS EVERYWHERE
There are a lot more cameras out there these days. Many homes have Ring cameras. Some phones have 3 little cameras all right next to each other for cooler pictures. And yes, there actually are cameras even smaller than a pinhead. You could say pinhead cameras aren’t common, but that is likely because quality and efficiency took priority over tininess.
KURZWEIL & GENERATIVE AI
2019 PREDICTIONS HAPPENING ONLY NOW
VIRTUAL ARTISTS
Kurzweil believed that in 2019, art of all sorts would be something an AI could generate. It may be a few years late, but with DALL-E, Midjourney, ChatGPT, and even AI-manufactured music, AI is now officially playing within all our major art forms.
AI ASSISTANTS
He envisioned that, as of 2019, computers would be close to a human-like assistant for us, helping us deal with our professional and daily lives all the time. He even saw lawyers using AIs as researchers or doctors consulting with them for advice. Sure, this may not be happening today, but Large Language Models like ChatGPT could become these virtual assistants soon.
AI EDUCATORS
Kurzweil saw us using computer-simulated teachers and tutors around 2019. And some think generative AI may reshape how we learn. However, if we discovered anything from the COVID educational experiment, human interaction is still key to academic and mental growth, especially for youth. So how much Kurzweil is right about AI becoming our main educators still seems in doubt for the near future.
WHAT KURZWEIL GOT WRONG
FROM “KINDA OFF” TO “BIG MISS”
COMPUTERS SOUND FULLY HUMAN
This is a hard call as there are some well-doctored deep fakes out there. Furthermore, we all have access to some very advanced voice generators, such as WellSaid Labs and soon Microsoft’s VALL-E, which are difficult to discern as artificial voices in short bursts. However, “fully human” is a high bar and if we listen long enough to these voices, we can still detect a tinge of automation, and this is a milestone Kurzweil thought we’d surpass by 2009.
AI-ROBOTIC DISABILITY ASSISTANCE
Kurzweil envisioned a 2019 where people with disabilities had very helpful computerized assists. Someone with a spinal cord injury could walk upstairs, blind people could wear special glasses that helped them interpret the world around them, and the deaf could see words spoken to them as people were talking. While we see impressive advancements of this type of tech on YouTube now (such as exoskeletons that help people walk or sunglasses for the deaf), they aren’t commonplace unfortunately.
BOOKS WILL BECOME OBSOLETE
Print isn’t dead. In a 2022 study, it was revealed that people still prefer paper books (65%) over E-Books (21%), and these purchases of tangible books vastly outnumber that of digital books too. So while devices like DVD Players, Tivos, non-phone cameras, and other recent tech are quickly fading, it may take a lot for simple words on paper to ever disappear. Do we need all these physical books, probably not, but it is our original form of media, and it may be more ingrained in humanity than we realize.
AI TAKES THE STEERING WHEEL
Kurzweil was not alone in this prediction. Many futurists felt that computers would be doing most of the driving by now. True, some modern cars can drive themselves in limited capacities. My Honda, for instance, knows how to stay in its own highway lane and can make speed adjustments when put on cruise control.
There are even arguments that autonomous vehicles are safer than human-driven ones.
Despite the impressive features, the technology keeps running into roadblocks. One CNN Business article put it this way back in November 2022: “The technology could do impressive things but mastering all the situations we face as human drivers is tough.” Even the CEO of Waymo LLC (formerly known as the Google Self-Driving Car Project) has said it will be decades before the roads are taken over by AI.
VR EYEWEAR IS EVERYWHERE
Kurzweil had this whole vision that by this time we’d be wearing 3-D VR glasses or contact lenses. And this eyewear was going to send digital displays right into our retina. He brought up these VR glasses many times in his predictions, describing a world where wherever we looked, we could do our emails, search the web, watch TV shows, play augmented-reality video games, and even interact with virtual assistants. Everything we’d see would be half reality and half computerized visualizations.
Today the technologies for “Virtual Retinal Displays” and “projections” do exist, but not in the way Kurzweil foresaw. What’s more, the idea of computerized eyewear hasn’t become a tech people wanted, let alone needed.
AI-ROBOTIC HOUSEKEEPERS
Kurzweil expected us to have a lot more than Roombas in 2023. He thought that we’d have one or more housekeeping robots in every home. To the relief of my two dogs, Roombas and their buddies haven’t caught on… unless you count a plain old dishwasher as a robot, in which case I’d argue it is the most amazing robot ever. How it cleans everything I cram in there always blows my mind.
VR CAN BE FELT, NOT JUST SEEN AND HEARD
Virtual Reality really only has audio and visual components at the moment. Kurzweil theorized that the technology would have gone a step further by now, in which we could feel the virtual environment by wearing special gloves and suits. This is called “haptic VR”, and while there have been some advancements in this vein, haptic technology is far from mainstream and it has a lot of challenges if it wants to match the audio-visuals we already experience in VR.
AI-DRIVEN MEDICINE
Probably Kurzweil’s most disappointing misses came in the field of medicine. He had some high hopes that by 2019 AI would help us “largely overcome the major diseases that kill 95 percent of us in the developed world” (predicted in 2006) and that life expectancy would be over 100.
A lot of this hope came from the idea that “bioengineered treatments for cancer and heart disease [would] have greatly reduced the mortality from these diseases.” Clearly, we are far short on all these marks.
HUMANS HAVE RELATIONSHIPS WITH AI
Kurzweil probably wasn’t thinking that we’d be falling in love with our computers like in the film Her (with Joaquin Phoenix), but he did believe that relationships with AI would be much more prevalent at this point in time.
He predicted that come 2019, we would verbally communicate with simulated people in full-blown back-and-forth conversations. He even thought there would be a visual element to these discussions, like head nods and gestures.
On top of that, we would now be thinking of our digital counterparts as advisors, friends, and yes, some of us would even believe that our computers possess a human-level of consciousness. At present, I’d say we are far from that relationship stage, and it’s hard to know when, if ever, we will get there.
MAJOR TAKEAWAYS
LESSONS FROM KURZWEIL’S PREDICTIONS
So Kurzweil got a lot wrong. In fact, in 2019, Stuart Armstrong of The Future of Humanity Institute assessed that Kurzweil had a 42% accuracy rate, and that was one of the more favorable grades. Even with his big misses, it does still feel like Kurzweil’s prognostications are moving in the directions he anticipated, though it often looks like he may be off by a decade or so.
In that respect, what’s cool is that he has given us a way to measure our progress. We can use these milestones to see how quickly we are trending toward that singularity event, or if it seems like progress has stalled out.
In a way, he has also given us something maybe even more important, at least for the near future. He has given us perspective on tech predictions and why they don’t always pan out.
If you look at what predictions didn’t come to fruition, I think you’ll see that a lot of the shortcomings may have had little to do with the advancements themselves, and more to do with human nature. Think about self-driving cars, computerized glasses, or even simulated people to interact with. The tech is more than halfway there, but that half doesn’t seem to make us that excited about the other half. If we’re honest, VR is completely here, but that’s not catching on either. It makes you wonder if there is a lack of need and want for certain technology, which will ultimately prevent them from becoming significant parts of our lives.
On the other hand, if we look at some of the biggest tech influences on society over the last 20 years, such as Amazon or the iPhone, I think we recognize that these were more than just cool ideas with big corporate pushes behind them. Those innovations were intuitive and made consumers’ lives a lot easier.
So an economist might argue that supply is not enough. There needs to be a demand, and for a rapid transformation, there needs to be a lot of that demand.
Thus, as we look at ChatGPT in the context of Kurzweil and his overestimations, we may learn that AIs don’t just need to do something amazing. They must expose and/or generate a great need in our lives.
The factors pushing for that will be of course convenience, efficiency, affordability, and intuitive integration. Smartphones and the Cloud arguably met these criteria.
But when we look at the misses, such as Kurzweil’s prediction regarding VR glasses or VR in general, maybe it is the “intuitive integration” that proves to be the most challenging hurdle. I think we find that just because a computer can do a cool trick, maybe we don’t like it when something becomes that easy or automated. Maybe there are things we need to just do ourselves.
It could be even more abstract than that. Take physical books for instance. We still prefer them over E-books, which are always conveniently right there on our phones. And then there are online interactions — it is easier to text with someone or do a zoom call. But we also learned from COVID isolationism that in-person interactions are better for us when it comes to our mental well-being or academic development.
So maybe a similar phenomenon will be prevalent when it comes to generative AI. As amazing as ChatGPT’s generated texts are or will be, we may still need to know that the communications we receive and give are from actual human beings.
I hope to explore these ideas more in future posts, but let me leave you with a few questions that I will be thinking about.
What is important to us when it comes to artwork, news articles, emails we get, or advertisements? Do we care whether or not there is a flesh-and-blood creator behind these things? Do we feel a need to be the ones conveying our own ideas?
_____
Be sure to check out my other posts, as I try to do little predictions of my own regarding the future of tech. Although, unlike Kurzweil, I hope to only look ahead maybe 20 months, not 20 years.
