Encode GPT

My friend's experience on the subway platform illustrates the impact of the new wave of large language models (LLMs) on the job market. OpenAI's GPT-4 and GPT-3. 5 models have scored above 80 percent on 11 AP and SAT exams, 77 percent on "Advanced Sommelier (theory knowledge)", and 90 percent on the Uniform Bar Exam, the national test to become a lawyer. This has upended the traditional wisdom that lower-wage, lower-skill jobs would be hardest hit by automation, as these models are now capable of performing tasks that were previously thought to require human expertise.

A few months ago, I was waiting for the subway with a friend, a professional editor, who had never used a large language model (LLM). Standing on the platform, she told me about an article she’d been working on. ChatGPT had come out six weeks earlier, and I input her summary into it on my phone and showed her the result. I’d been following OpenAI’s transformer-driven models since 2019 and had forgotten the effect they can have on first exposure. My friend couldn’t take her eyes off the little gray box as the article came out, line by line. It took me a minute to register the shock on her face. On the train, she said, only half-joking, “I’m going to be unemployed by the end of the year.”As wave after wave of new AI capabilities have hit over the past few months, I’ve been thinking of my friend and her place in the world that’s unfolding. When GPT-4 came out in March, OpenAI’s press release included a chart of its scores on various standardized tests. The much-hyped new model scored above 80 percent on 11 AP and SAT exams, 77 percent on “Advanced Sommelier (theory knowledge),” and—most buzzed-about on Twitter—90 percent on the Uniform Bar Exam, the national test to become a lawyer. OpenAI’s previous model, GPT-3.5 (which powered ChatGPT when it debuted), had already passed the US Medical Licensing Examination, earning a grade that, were it human, would qualify it to become a doctor.Results like these seem to validate a <a target="_blank" rel="noopener noreferrer nofollow" href="https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3482150">2019 paper</a> by then Stanford PhD candidate Michael Webb. Though entirely speculative at the time, it upended received wisdom about who would win and who would lose as a result of AI-driven automation. Before Webb’s report, studies by <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf">Oxford</a> and <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages">McKinsey</a> had predicted that lower-wage, lower-skill jobs would be hardest hit, as indeed they have been throughout the entire history of automation going back to the steam-powered weaving loom.The LLM era has changed all that. Now, the conventional wisdom—replicated and extended in <a target="_blank" rel="noopener noreferrer nofollow" href="https://arxiv.org/abs/2303.10130">a recent paper</a> by researchers at OpenAI—is that higher-paid jobs and creative jobs (including mathematicians, tax preparers, quants, writers, and web designers, to name a few) are the mostly highly exposed to automation (100 percent exposure for the professions just listed). This has an interesting side effect, since, as Webb’s study spelled out, white-collar work in the US is <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.brookings.edu/research/what-jobs-are-affected-by-ai-better-paid-better-educated-workers-face-the-most-exposure/">disproportionately done</a> by the most privileged: men, white people and Asian Americans, people in their prime working years (25-54), and people living in rich coastal cities. Many in these demographics have had it easy for a long time, but it’s looking like the AI revolution will be a bumpy ride for them.I spoke with four economists for this piece, and, though they offered good reasons to think AI won’t “take all the jobs”—indeed, as in previous waves of automation, the economy is likely to grow—none denied that some jobs will be lost. They didn’t know exactly how many, and neither do I. But what I do know is that we've never had a wave of automation in which white-collar workers are uniquely vulnerable, and we should therefore expect this one to play out differently.The crux of the difference lies in the relationship that blue-collar and white-collar workers have with work. According to <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.researchgate.net/publication/308042364_Meaningful_work_differences_among_blue-_pink-_and_white-collar_occupations">one study</a>, white-collar workers tend to feel that they “express” their “full potential” more at work than blue-collar workers do; they also experience higher levels of “developing inner self” at work. According to <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.pomsmeetings.org/confpapers/043/043-1565.pdf">another study</a>, white-collar workers valued “interesting work (nature of the work), achievement, and appreciation of the work done (recognition)”—in contrast to blue-collar workers, whose motivators were “receiving salary, working condition, peer relations, and job security.” (And even more than other groups, men derive their self-worth from achievement and feeling useful. A dramatic illustration of this was <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4611172/">a study</a> of language used by suicidal men, which showed that being considered useful was core to men’s well-being, and its absence was devastating. Being rendered useless by a bot will have disproportionately bad emotional effects for a man.)“Nature of the work” is one way of saying that white-collar workers care about the tasks we do. Being “recognized” and “appreciated” for “achievement” in these tasks is important to us; it is how we “express” our “full potential.” In other words, large pieces of our emotional lives and social selves are hooked into the tasks we do for work. What happens when AI does those tasks better?At the furthest edge of white-collar work is a species of task where competency is so admired that it becomes a sport or an art, and competency is rewarded by society with status and esteem, over and above financial compensation. This is the category of games of logic and art. Our shock at the new wave of AI models like ChatGPT and Midjourney comes from their proficiency at the artier, more creative tasks like writing and illustration. But the more strictly logical sports of chess and Go have long since been conquered by previous waves of AI, and so it may be instructive for the left-brainers to look in on how the right-brainers have handled usurption, emotionally and practically.Go is generally considered humanity’s most compex game. In 2016, DeepMind’s AlphaGo beat two of its highest-ranked players. Lee Sedol, a Korean prodigy and the second-best player in the world at the time, took it the hardest. He became depressed, and a couple years after the match he retired from the game, citing AlphaGo. “Even if I become the number one, there is an entity that cannot be defeated,” he said.Fan Hui, the European champion but a rung below the world leaders, took it better. He was initially shocked and humbled by his defeat, and indeed tried to forget the game altogether. “I want to try to forget Go, but it’s impossible, because all the things I’ve learned in my life is with Go,” he said in AlphaGo, the 2017 documentary released by DeepMind about Lee and Fan. The game, he said, is like looking in a mirror. “I see Go; I also see myself. For me Go is real life.” Later, however, he joined DeepMind—the architect of his defeat—and helped improve its model’s capabilities. Basically, he couldn’t beat ’em, so he joined ’em.These differences seem illuminating, and it’s hard not to look for lessons. I can’t help but think that Lee’s higher ranking actually made him more vulnerable to an existential crisis, because he had more to lose. Fan was disappointed, but Lee lost very publicly in front of millions of Korean viewers. It’s hard to bounce back from that, and perhaps harder than it was for Fan to pivot.I caught up with my editor friend again recently—three months after her first exposure to ChatGPT. She seemed more concerned than ever. “I just think it’s going to be a hard fall,” she said. She felt the younger, more technically adept nipping at her heels and was worried she hadn’t been brought up to be resilient enough for this kind of challenge. I tried to offer hope in the form of a story that Gregory Clark, a professor emeritus at UC Davis, told me about aristocratic land owners during the Industrial Revolution. Tenant farmers abandoning the country to follow better wages into factories in the city caused the value of the aristocrats’ farmland to drop, causing massive losses for the aristocracy. The smart aristocrats, though, said Clark—the ones who could adapt—simply followed the farmers into the cities and became urban landlords.My friend was only partly sold. What was the equivalent now, for her?That’s when I remembered a third Go champion who played AlphaGo but wasn’t included in the documentary. This is Ke Jie. In 2017, months after the Lee match, he was 19 years old and the best player in the world, having beaten Lee in three consecutive championships. Like Fan and Lee, Ke also lost to AlphaGo, after which AlphaGo had no human left to beat.But Ke’s reaction is, I think, the most interesting and also the most hopeful. Pre-AlphaGo, Ke, a teenager of world-class abilities, was also a world-class brat, famous for bucking Go’s culture of humility. When Ke challenged Lee to a match, for example, he posted a video of himself as a boxer beating up Lee and ostentatiously bragged and baited his opponents.In the aftermath of Ke’s defeat by DeepMind’s AI, however, he underwent a remarkable change. On TV appearances since then, he has affected a stance of irony, playfulness, and humility, becoming a much loved crowd-pleaser along the way. Again, looking for lessons, I can’t help but notice Ke’s extreme youth—15 years younger than Lee, 16 years younger than Fan—and wonder if he had less invested in a particular way of valuing and understanding himself. Perhaps he was therefore better able to change how he related to the world on a fundamental level.Important to this story, too, is that, unlike Fan, whose pivot to temporary AI research consultant could be seen as a demotion from European Go champion, Ke’s pivot allowed him to remain at the top of the game.The pivot from “best player in the world at humanity’s most logically complex game” to “comedian” is pretty dramatic, though, and I think the magnitude of that flip reflects the profundity of the changes coming down the pipe. And if Ke Jie has to do that, what does that mean for the rest of us? My hunch is that economic concerns will dominate in the coming years, but assuming that’s solved, where will status reemerge if the core competencies of art, design, science, law, medicine, and engineering are swallowed by GPT-7?Webb himself thought the human niche would become something closer to judgment, “where the point is that it’s a human making the decision.” For a judge or a politician or a newspaper editor, for example, “we know we could get the AI to do it for us—we could ask it to tell us what to do—but we’d rather have a human do it.”Again, the vanguard of Go and chess—“solved” by AI two decades earlier—offer us tea leaves to divine if we choose to read them. In these worlds, Ke Jie is not the only high-status genius to pivot as he did; Magnus Carlsen, the world’s best chess player, has in recent years become known for “interesting” gameplay in response to AI creating an indisputable hierarchy of opening moves. Even more heretical, players at much lower skill levels are beginning to overtake the old masters in popularity: The personable and attractive Botez sisters are <a target="_blank" rel="noopener noreferrer nofollow" href="https://www.twitchmetrics.net/channels/viewership?game=Chess">the second-most-streamed</a> chess players while having ELO ratings nowhere near the world’s best. And Zhan Ying, a Chinese Go player at a skill level considerably below Ke Jie’s, recently dethroned him, briefly, as the most-watched Go player in the world.If this trend is any indication, we should expect to see softer skills—humor, presence, personality—become the game. In this light, we may already be halfway there without quite realizing it: Perhaps the future belongs to the influencer.

 A few months ago, I was waiting for the subway with a friend, a professional editor, who had never used a large language model (LLM). Standing on the platform, she told me about an article she’d been working on. ChatGPT had come out six weeks earlier, and I input her summary into it on my phone and showed her the result. I’d been following OpenAI’s transformer-driven models since 2019 and had forgotten the effect they can have on first exposure. My friend couldn’t take her eyes off the little gray box as the article came out, line by line. It took me a minute to register the shock on her face. On the train, she said, only half-joking, “I’m going to be unemployed by the end of the year.” As wave after wave of new AI capabilities have hit over the past few months, I’ve been thinking of my friend and her place in the world that’s unfolding. When GPT-4 came out in March, OpenAI’s press release included a chart of its scores on various standardized tests. The much-hyped new model scored above 80 percent on 11 AP and SAT exams, 77 percent on “Advanced Sommelier (theory knowledge),” and—most buzzed-about on Twitter—90 percent on the Uniform Bar Exam, the national test to become a lawyer. OpenAI’s previous model, GPT-3.5 (which powered ChatGPT when it debuted), had already passed the US Medical Licensing Examination, earning a grade that, were it human, would qualify it to become a doctor. Results like these seem to validate a 2019 paper by then Stanford PhD candidate Michael Webb. Though entirely speculative at the time, it upended received wisdom about who would win and who would lose as a result of AI-driven automation. Before Webb’s report, studies by Oxford and McKinsey had predicted that lower-wage, lower-skill jobs would be hardest hit, as indeed they have been throughout the entire history of automation going back to the steam-powered weaving loom. The LLM era has changed all that. Now, the conventional wisdom—replicated and extended in a recent paper by researchers at OpenAI—is that higher-paid jobs and creative jobs (including mathematicians, tax preparers, quants, writers, and web designers, to name a few) are the mostly highly exposed to automation (100 percent exposure for the professions just listed). This has an interesting side effect, since, as Webb’s study spelled out, white-collar work in the US is disproportionately done by the most privileged: men, white people and Asian Americans, people in their prime working years (25-54), and people living in rich coastal cities. Many in these demographics have had it easy for a long time, but it’s looking like the AI revolution will be a bumpy ride for them. I spoke with four economists for this piece, and, though they offered good reasons to think AI won’t “take all the jobs”—indeed, as in previous waves of automation, the economy is likely to grow—none denied that some jobs will be lost. They didn’t know exactly how many, and neither do I. But what I do know is that we've never had a wave of automation in which white-collar workers are uniquely vulnerable, and we should therefore expect this one to play out differently. The crux of the difference lies in the relationship that blue-collar and white-collar workers have with work. According to one study , white-collar workers tend to feel that they “express” their “full potential” more at work than blue-collar workers do; they also experience higher levels of “developing inner self” at work. According to another study , white-collar workers valued “interesting work (nature of the work), achievement, and appreciation of the work done (recognition)”—in contrast to blue-collar workers, whose motivators were “receiving salary, working condition, peer relations, and job security.” (And even more than other groups, men derive their self-worth from achievement and feeling useful. A dramatic illustration of this was a study of language used by suicidal men, which showed that being considered useful was core to men’s well-being, and its absence was devastating. Being rendered useless by a bot will have disproportionately bad emotional effects for a man.) “Nature of the work” is one way of saying that white-collar workers care about the tasks we do. Being “recognized” and “appreciated” for “achievement” in these tasks is important to us; it is how we “express” our “full potential.” In other words, large pieces of our emotional lives and social selves are hooked into the tasks we do for work. What happens when AI does those tasks better? At the furthest edge of white-collar work is a species of task where competency is so admired that it becomes a sport or an art, and competency is rewarded by society with status and esteem, over and above financial compensation. This is the category of games of logic and art. Our shock at the new wave of AI models like ChatGPT and Midjourney comes from their proficiency at the artier, more creative tasks like writing and illustration. But the more strictly logical sports of chess and Go have long since been conquered by previous waves of AI, and so it may be instructive for the left-brainers to look in on how the right-brainers have handled usurption, emotionally and practically. Go is generally considered humanity’s most compex game. In 2016, DeepMind’s AlphaGo beat two of its highest-ranked players. Lee Sedol, a Korean prodigy and the second-best player in the world at the time, took it the hardest. He became depressed, and a couple years after the match he retired from the game, citing AlphaGo. “Even if I become the number one, there is an entity that cannot be defeated,” he said. Fan Hui, the European champion but a rung below the world leaders, took it better. He was initially shocked and humbled by his defeat, and indeed tried to forget the game altogether. “I want to try to forget Go, but it’s impossible, because all the things I’ve learned in my life is with Go,” he said in AlphaGo, the 2017 documentary released by DeepMind about Lee and Fan. The game, he said, is like looking in a mirror. “I see Go; I also see myself. For me Go is real life.” Later, however, he joined DeepMind—the architect of his defeat—and helped improve its model’s capabilities. Basically, he couldn’t beat ’em, so he joined ’em. These differences seem illuminating, and it’s hard not to look for lessons. I can’t help but think that Lee’s higher ranking actually made him more vulnerable to an existential crisis, because he had more to lose. Fan was disappointed, but Lee lost very publicly in front of millions of Korean viewers. It’s hard to bounce back from that, and perhaps harder than it was for Fan to pivot. I caught up with my editor friend again recently—three months after her first exposure to ChatGPT. She seemed more concerned than ever. “I just think it’s going to be a hard fall,” she said. She felt the younger, more technically adept nipping at her heels and was worried she hadn’t been brought up to be resilient enough for this kind of challenge. I tried to offer hope in the form of a story that Gregory Clark, a professor emeritus at UC Davis, told me about aristocratic land owners during the Industrial Revolution. Tenant farmers abandoning the country to follow better wages into factories in the city caused the value of the aristocrats’ farmland to drop, causing massive losses for the aristocracy. The smart aristocrats, though, said Clark—the ones who could adapt—simply followed the farmers into the cities and became urban landlords. My friend was only partly sold. What was the equivalent now, for her? That’s when I remembered a third Go champion who played AlphaGo but wasn’t included in the documentary. This is Ke Jie. In 2017, months after the Lee match, he was 19 years old and the best player in the world, having beaten Lee in three consecutive championships. Like Fan and Lee, Ke also lost to AlphaGo, after which AlphaGo had no human left to beat. But Ke’s reaction is, I think, the most interesting and also the most hopeful. Pre-AlphaGo, Ke, a teenager of world-class abilities, was also a world-class brat, famous for bucking Go’s culture of humility. When Ke challenged Lee to a match, for example, he posted a video of himself as a boxer beating up Lee and ostentatiously bragged and baited his opponents. In the aftermath of Ke’s defeat by DeepMind’s AI, however, he underwent a remarkable change. On TV appearances since then, he has affected a stance of irony, playfulness, and humility, becoming a much loved crowd-pleaser along the way. Again, looking for lessons, I can’t help but notice Ke’s extreme youth—15 years younger than Lee, 16 years younger than Fan—and wonder if he had less invested in a particular way of valuing and understanding himself. Perhaps he was therefore better able to change how he related to the world on a fundamental level. Important to this story, too, is that, unlike Fan, whose pivot to temporary AI research consultant could be seen as a demotion from European Go champion, Ke’s pivot allowed him to remain at the top of the game. The pivot from “best player in the world at humanity’s most logically complex game” to “comedian” is pretty dramatic, though, and I think the magnitude of that flip reflects the profundity of the changes coming down the pipe. And if Ke Jie has to do that, what does that mean for the rest of us? My hunch is that economic concerns will dominate in the coming years, but assuming that’s solved, where will status reemerge if the core competencies of art, design, science, law, medicine, and engineering are swallowed by GPT-7? Webb himself thought the human niche would become something closer to judgment, “where the point is that it’s a human making the decision.” For a judge or a politician or a newspaper editor, for example, “we know we could get the AI to do it for us—we could ask it to tell us what to do—but we’d rather have a human do it.” Again, the vanguard of Go and chess—“solved” by AI two decades earlier—offer us tea leaves to divine if we choose to read them. In these worlds, Ke Jie is not the only high-status genius to pivot as he did; Magnus Carlsen, the world’s best chess player, has in recent years become known for “interesting” gameplay in response to AI creating an indisputable hierarchy of opening moves. Even more heretical, players at much lower skill levels are beginning to overtake the old masters in popularity: The personable and attractive Botez sisters are the second-most-streamed chess players while having ELO ratings nowhere near the world’s best. And Zhan Ying, a Chinese Go player at a skill level considerably below Ke Jie’s, recently dethroned him, briefly, as the most-watched Go player in the world. If this trend is any indication, we should expect to see softer skills—humor, presence, personality—become the game. In this light, we may already be halfway there without quite realizing it: Perhaps the future belongs to the influencer. My friend's experience on the subway platform illustrates the impact of the new wave of large language models (LLMs) on the job market. OpenAI's GPT-4 and GPT-3.5 models have scored above 80 percent on 11 AP and SAT exams, 77 percent on "Advanced Sommelier (theory knowledge)", and 90 percent on the Uniform Bar Exam, the national test to become a lawyer. This has upended the traditional wisdom that lower-wage, lower-skill jobs would be hardest hit by automation, as these models are now capable of performing tasks that were previously thought to require human expertise. 

My friend's experience on the subway platform illustrates the impact of the new wave of large language models (LLMs) on the job market. OpenAI's GPT-4 and GPT-3.5 models have scored above 80 percent on 11 AP and SAT exams, 77 percent on "Advanced Sommelier (theory knowledge)", and 90 percent on the Uniform Bar Exam, the national test to become a lawyer. This has upended the traditional wisdom that lower-wage, lower-skill jobs would be hardest hit by automation, as these models are now capable of performing tasks that were previously thought to require human expertise.

Who Will You Be After ChatGPT Takes Your Job?

https://us-prod-content.primeobjects.net/v1/

DouHub Team

types

All partners

All competitors

All customers/clients

All investors

All service/product providers

All knowledge/idea providers

Organization Administrator

Content Administrator

User Administrator

License Administrator

User

Help you with product management, marketing, R&D, and operations to ensure a successful SaaS business from Zero to One.

Who Will You Be After ChatGPT Takes Your Job?

More Readings