From Kasparov’s chess career to Steve Jobs’ bicycle, we already know how to approach these new AI tools. The real question: Will you master them before your competition does? https://t.co/WdrN5CT3yn
— PCMag (@PCMag) February 13, 2023
This article is a reprint. You can read the original at PCGamer.
By Brian Westover
“By now, you have no doubt heard of ChatGPT, the artificially intelligent (AI) chatbot that just about everyone online is talking about. Unless you’ve been living under a rock, you’re probably one of the millions of people already trying out the new tool, experimenting with its capabilities, playing with its responses, and finding all the fun and strange ways that generative AI can frustrate you.
Even if you haven’t used ChatGPT, you’ve surely read about it. A lot of digital ink has been spilled about the new chat bot, even though it’s only been available to the public for about three months. Since then, everyone from workers to students to cybercriminals have found new ways to use and abuse ChatGPT. The majority of coverage is a strange mix of over-inflated hype and profound pessimism, insisting that AI is coming for our jobs (or letting you make money without a job), and that no one is safe.
Both of these views are missing the point.
It makes sense that such a profoundly disruptive tool would generate such a dramatic response. The hype cycle(Opens in a new window) is real, and the temptation to cash in on a trend is never something passed up by a certain segment. Just try searching for information about ChatGPT on YouTube, and you’ll be inundated with videos promising that you can make millions with ChatGPT, without working at all! They’re also looking for easy money—from you, and any other mark they can con into buying the book, course, or product they’re selling. As ever, the wise move is to ignore these con artists.
Doomsayers are also common whenever any disruptive technology arrives. Ironically, these naysayers are nearly robotic in their frowny forecasting. They all seem to have two things in common: First, they are always quick to downplay what’s happening, or to overemphasize the negatives. Second, and perhaps more important, they are almost always wrong.
In the past, this same negative inclination led experts to pooh-pooh the idea that the internet(Opens in a new window) would ever be important in our lives, or that the iPhone was more than a iPod that could make calls. There’s always a Blockbuster ready to say no(Opens in a new window) when Netflix is available to buy.
However, this opinion isn’t meant to be over-inflated hype or unimpressed cynicism. Instead, I prefer to be skeptical of the skeptics, and ignore the hype machine entirely. I want to inject some reality into how people think about the coming wave of AI tools.
I’m here to talk about bicycles—but, I will start by talking about chess.
In 1996, chess world champion and Grandmaster Garry Kasparov faced off against Deep Blue, an advanced chess playing computer built by a team at IBM. He won, just as humans had been beating computers for decades before. That’s why you likely don’t know about the 1996 match—instead, you more likely know about 1997.
In a rematch with a re-tooled Deep Blue, Kasparov made real, notable history, as the first top-ranked human player on Earth to lose to a machine(Opens in a new window). It was a pivotal moment for the progression of computational thinking, as it proved that the statistical brute force a computer can leverage was capable of defeating the very best that a skilled human could do on their own.
At the time, people experienced a similar cycle of hype-filled and apocalyptic news coverage. As journalists and pundits feared that the game of chess was over, tech optimists cheered that we could turn over all of our thinking to the coming wave of intelligent machines. The cold embrace of pure logic would change humanity forever.
Yet chess didn’t disappear. Kasparov’s reign as world champion continued for several years until his retirement, and chess may be more popular now than ever before. And this, despite the fact that many chess-playing phone apps today can outperform anything Deep Blue was capable of so many years ago.
Apple cofounder Steve Jobs once famously called the personal computer a “bicycle for the mind(Opens in a new window),” explaining that while humans are outpaced in speed and energy efficiency by any number of animals in the natural world, a human on a bicycle handily beats them all. In Jobs’ thinking, and in several subsequent Apple commercials, the message was clear: Humans are tool users, and better tools make us better humans.
While a decade of tech journalism has taught me to question much of the marketing around any tech product, that central concept has stuck with me all my life. Technology augments and enhances what humans are capable of.
Yes, the quirks of technological progress, both the expanded capabilities and the practical and conceptual limitations of new tools, shape human society and culture in turn. I’m not saying we ignore the potential misuses of technology, or fail to consider the problems inherent in any new tool. I am saying that smarter tools enable us to progress above and beyond the limitations of the past. And, if we are smart in how we apply them, they can help us not “merely” at a grand societal level, but on a personal level.
AI presents us with a new, more capable bicycle. We’re ditching the huge front wheel of the penny-farthing and adding gear shifts, disk brakes, shock-absorbing forks and frames, and GPS directions. Like the real bicycle, these tools are rapidly evolving. However, the human still has to pedal and steer the bike, and AI tools have to be powered by human thinking and human ingenuity. The bike just helps the human move faster.
Any tool, properly applied, increases human capability—but more powerful tools will lengthen our reach and increase our impact even when we use them poorly, and that’s what’s scary. Smarter tools demand that we be smart in how we use them.
We’re already getting our first taste of how the next generation of tools can be used and misused. College kids are trying to turn in AI-generated papers. Script kiddies are getting powerful coding tools to make malware. Short-sighted executives are pushing AI-generated media, oblivious to the fact that it may be full of inaccurate info, or even plagiarized content. And, my personal favorite: Somebody posted a ChatGPT-generated job description(Opens in a new window) without even reading it first.
AI presents us with a new, more capable bicycle….however, the human still has to pedal and steer the bike.
But, if you want to make the most of a new tool, the answer isn’t to run and hide from it. Let’s go back to talking about chess.
In 1998, just a year after his defeat to Deep Blue, Garry Kasparov was again at the center of a seismic change in the chess world. Instead of abandoning the game he loved, or capitulating to the idea of unbeatable machines, he chose to adapt by embracing new technology. He created a new version of the game called Advanced Chess(Opens in a new window).
This new approach would pair a human player with one or more chess computers, working as a team. Borrowing an image from mythology, the human intellect steering the computer workhorse was called a “centaur team.” And within a few years, centaur teams made up of amateur players and midrange computers were beating both human grandmasters and top-of-the-line chess computers.
The answer to the questions posed by today’s advancements has already been found. Human intellect and creativity, paired with powerful tools, is the winning combination. It always has been.
Computers are tools, and as powerful and as fascinating as the latest AI-driven chatbots appear, they are still tools. They need a human to steer their capabilities, to evaluate the output, and to apply them intelligently. Somebody has to ride the bike.
The problems we’re going to run into in the near future are going to come from forgetting that relationship. We will see businesses harmed by foolishly trying to replace human writers with generative content, when the answer is human writers enabled by generative content. We’re going to see human artists fret over the end of artwork as image generators make it cheap and easy to create impressive visuals from a simple text prompt. But art isn’t going anywhere, and the essentials of creative composition, finding ways to surprise an audience, and having the skill to recognize what’s good and bad, and fine-tune it accordingly, those will all still rest in the hands of humans.
Just as chess didn’t end when Kasparov lost to Deep Blue, math didn’t end with the advent of the calculator. People feared that having a personal device that can handle all of the complex calculations that students were supposed to do would destroy the study of math. It didn’t.
Human intellect and creativity, paired with powerful tools, is the winning combination. It always has been.
Instead, mathematicians set down their slide rules and picked up their calculators, and they were able to advance the study of mathematics beyond what could be done previously. Students still study math. Mathematicians are still needed in the world, perhaps more than ever. And the humble calculator became the personal computer, which became the smartphone, which became whatever device you’re reading this on right now (probably a phone).
Once powerful machines come along, humanity doesn’t go away. Some Luddites may smash the looms in fear for their jobs—others embrace it.
With tools like ChatGPT available to the public only within the last few months, and other generative technologies, like Mid Journey and Stable Diffusion, arriving only last year, we are still in the very early days of an AI revolution. But make no mistake: It is a revolution, a genuine step change in the progress of technology.
I’ll confess: I use ChatGPT. A lot. Every day.
But first, to put readers (and my editor) at ease: I don’t use it for writing the articles and reviews you read under my byline—because that’s my job. Anything published under my byline is 100% written by a human. And not just any human, but me, in particular.
However, AI can do many, many things better and faster than I can, freeing me up to focus on the parts of my job that I’m uniquely capable of doing. So, I use ChatGPT and other AI tools to enhance my work processes. I’ve used it to build custom tools for data gathering and analysis that I use in my job. ChatGPT is also excellent at summarizing complex information, so that I can learn more technical information more quickly. I use it for brainstorming, I use it to fine-tune my planning, and I use it to enhance my life outside of work.
Ever wanted someone to suggest a recipe, turn that into a shopping list, and then organize the shopping list by section of the grocery store? ChatGPT can do that.
The question is, are we learning to use these new tools? It’s easy to hide from the new. There is some bliss in the ignorance of what a tool can do, and what opportunities it opens up. But you shouldn’t be afraid of an AI taking your job. You should be far more worried about another human, using AI, because they might take your job. If they’re halfway capable with the new tools, they’ll do your job better than you do. After all, amateurs are beating grandmasters these days, and no matter how much exercise you do, you won’t outrun a human who’s on a bicycle.
In the long run, the centaur wins. And, at that point, you need to decide whether you’ll be the human at the front—or the one looking at the horse’s rear end.”