The AI Daily Brief: Artificial Intelligence News and Analysis - AI Doesn't Have to Automate Away All the Jobs

Episode Date: April 27, 2024

A reading and discussion inspired by https://www.washingtonpost.com/opinions/2024/03/19/artificial-intelligence-workers-regulation-musk/ ** Join NLW's May Cohort on Superintelligent. Use code nlwmay for 25% off your first month and to join the special learning group. https://besuper.ai/ ** Consensus 2024 is happening May 29-31 in Austin, Texas. This year marks the tenth annual Consensus, making it the largest and longest-running event dedicated to all sides of crypto, blockchain and Web3. Use code AIBREAKDOWN to get 15% off your pass at https://go.coindesk.com/43SWugo  ** ABOUT THE AI BREAKDOWN The AI Breakdown helps you understand the most important news and discussions in AI.  Subscribe to The AI Breakdown newsletter: https://theaibreakdown.beehiiv.com/subscribe Subscribe to The AI Breakdown on YouTube: https://www.youtube.com/@TheAIBreakdown Join the community: bit.ly/aibreakdown Learn more: http://breakdown.network/

Today on the AI Breakdown, we're talking about how AI could help end the dominance of the credentialed class. The AI breakdown is a daily podcast and video about the most important news and discussions in AI. Go to Breakdown.network for more information about our YouTube, our newsletter, and our Discord. Hello, friends. Happy weekend, and it being a weekend that means, of course, it is a long-reads episode. Today, we turn to a column from Eduardo Porter called, as I hinted, AI could help ending the dominance of the credentialed classes. It hits at a theme that I think is coming up more and more, and which is really, really important. Which, in a nutshell, is that the impact that AI has on workers, on the workplace, is not some predetermined thing, but something that we have some amount of control over. Let's read

the piece, and then we will come back and discuss it a little bit further. Eduardo writes, there is good reason to believe that artificial intelligence is coming for our jobs. That's, after all, what information technology has been doing for decades. Tech plutocrats apparently agree that human employment will become the stuff of folklore. They probably don't mind that all wealth in this world will accrue to the lucky few who own the AI patents. They seem willing to share some of it with regular people through some universal basic income scheme. Artificial general intelligence the end goal Silicon Valley technologists are striving for is defined in the charter of OpenAI, the outfit that developed ChatGPT and Dali as, quote, highly autonomous systems that

outperform humans at most economically valuable work. Elon Musk has concluded that, quote, there will come a point where no job is needed. But work as we know it doesn't have to end this way. The sense of inevitability about our technological future stems from a misunderstanding of how technologies emerge and become embedded in the economy. The process is not driven by laws of nature. It hinges on political decisions that can lead to vastly different outcomes. North Korea devoted its nuclear knowledge to weapons. Japan deployed it to produce electric power. So far, the march of computers into the workplace has furthered Musk's vision, automation powered by information technology, has spread to take over every task that could be decomposed into a defined set of precise instructions

to placing swaths of blue-collar workers. AI could feasibly push automation even farther, but it could also reshuffle opportunity in a different direction. As Duran Asimoglu, David Atoor, and Simon Johnson of MIT have pointed out, it could be deployed not to replace human workers but to augment them, enabling them to perform more complex tasks. Examples of this alternative abound. For instance, research by economists at Stanford University and MIT found that a generative AI conversational assistant substantially increased the productivity of customer support agents measured in cases resolved per hour and improved customer satisfaction. The gains were greatest among the most novice workers. On the other end of the wage spectrum, computer programmers have also been found to benefit

from AI help. In an experiment by researchers at Microsoft, GitHub and MIT, coders who use the AI tool GitHub co-pilot, completed programming tasks 56% faster. It's hard to understate how much human enhancement via AI tools could reshape society. The vast inequalities in income and opportunity that automation has given us so far, could be significantly reduced. Computers didn't just automate routine blue-collar work. By vastly expanding access to information, computers plugged into the internet empowered the expert white-collar workers who could profitably use it. Doctors, programmers, lawyers broadly, workers with college degrees reap the rewards, leaving behind a host of less-educated workers whose jobs have progressively been taken over by the machines. AI could narrow the opportunity

gap by helping lower-ranked workers take on decision-making tasks currently reserved for the dominant credentialed elites. Consider how old-school IT tools such as electronic medical records have allowed nurse practitioners to make more complex decisions encroaching on doctors' turf, much to the chagrin of the American Medical Association. Generative AI could take this further, allowing nurses and medical technicians to diagnose, prescribe courses of treatment, and channel patients to specialized care. As Autor put it, AI, if used well, can assist with restoring the middle-skill, middle-class heart of the U.S. labor market that has been hollowed out by automation and globalization. Can the deployment of artificial intelligence be nudged in this direction, it would require a different approach

to technology, both by policymakers and the corporate class. Asimoglu, Ator, and Johnson say, the human complementary approach is not likely to prevail based on current investments and corporate attitudes. Corporate managers today itch to bring in the robots, even when the machines often don't do much to improve productivity. Perhaps it's because they don't agitate to join unions? In any event, policy could rejigger corporate incentives. Changing the tax code, which vastly favors capital investment over hiring new workers, would level the playing field and temper incentives to automate. Asimoglu, Autor, and Johnson argue for giving unions a voice on how to deploy the new technologies in the workforce. Regulations could stop companies from

deploying untested AI in tasks like hiring or surveillance, areas where technology already is degrading and devaluing work. Incentives could be provided to change the research agenda, which today is acutely focused on automation to explore complementary uses of AI instead. This requires a conceptual challenge, a paradigm of technological development that has long focused on replacing human capabilities must seat its place to a mindset concentrated on how to expand human potential. The stakes are high, as Stanford's Eric Brynjolfson wrote, AI could well generate enormous wealth and vastly enhance our understanding of the world. But if technology continues to replace people, the endgame includes a working class of no economic or political power, unable to improve

its lot, left to live off whatever universal basic income the Silicon Valley plutocrats are willing to provide. Hello friends, quick note before we get back to the show, I'm so excited to share that Super Intelligent is now live. Super Intelligent is a platform for fast, fun, and super practical, useful AI learning. We have something like 300 video tutorials adding 30 to 50 each week, covering every topic in AI you can imagine

from LLMs to image generators to case studies, use cases, basically everything that tells you how to use AI and what to use it on in short, fast, four to seven minute tutorial videos, which are paired with step-by-step instructions that help you actually use these tools as well. well. It's $20 a month for unlimited access, and I would love to see you there. Check it out at

B-super.a.I. That's B-super.A.I. All right, breakers. Consensus 2024 marks the 10th gathering of the biggest event that's devoted to all sides of the crypto, blockchain, and Web3 ecosystems. Join pioneering fingers and builders as they delve into the future of Defy and explore game-changing tech, from AI to ZK Proofs and everything in between. The event is three days of jam-packed content, networking, and so much more. Some of the speakers at the event include Chris Dixon, the founder and managing partner at A16Z Crypto, Sergey Nazaroff, the co-founder of Chainlink, Kathy Wood, the CEO of Arc, Hester Purse, commissioner, of course, from the U.S.S. SEC, and Tom Emmer, Republican Majority Whip for the U.S. House of Representatives.

Visit Consensus24.coindex.com to learn more and save 15% on registration with the code breakdown. That is 15% on registration with the code breakdown. All right, back to NLW here. So there's a bunch that I think is really important and valuable about this conversation. Let's separate it perhaps into the general and then the specific. And actually, let's do it in reverse order. The specifics here, lots and lots of people would want to debate, right? Are unions, for example, the best way to make these sort of shifts? Is policy the best vehicle? Are different types of incentives? Perhaps in the tax code? Are these the ways that you could get companies, managers, the investment class, etc., to think differently about AI in its role in automation?

More broadly, though, hold aside any of those specifics. There is an important core idea here, which seems obvious, but actually isn't really considered very much, which is that we get to have a say, in other words, in how technology is applied to our lives. On the broadest level, there is an inexorable march towards more rather than less technology. That is the story of human history. However, how technology comes into and interacts with any given society is not just some inexorable march that we have no control over. I'm very encouraged that you have more people speaking up and talking about how artificial intelligence, which seems so obviously like a force for even more automation than we've ever had before, could also significantly augment human potential in different ways as well.

I want that conversation to be louder. In fact, I want the conversation to shift from binaries like blanket bands on the one hand or EACC and accelerationism on the other to this type of conversation where we're talking about specifics. Now, of course, there is a whole additional piece of this, which is the unpredictability of what comes on the other side of pure automation. If we really did get to these versions of abundant utopia where nobody's labor was needed anymore, what would society look like? I also find those conversations not necessarily super useful in the short term, and I'm much more interested in realities that are a little bit closer to the here and now.

As I've shared before on this show, my base case is that as basically every new technology has done before it, AI will create entire sets of opportunities, entirely new industries, that we literally can't imagine today. And the challenge of trying to think about policy or our approach to this industry is that we're imagining it in the context of it just doing the same stuff we're doing now, just faster, cheaper, or better. The way the technology makes things better tends to be on the other side of that base-level automation, where new possibilities become unlocked. My general belief is that a world in which there is more creation possible for cheaper means we'll just make more stuff. I don't think that everyone's going to stop working. I don't even really think

that people are going to work much less. I think instead, they're going to be able to do way more in the time that they have. In other words, if each coder in 10 years can do 10 times the amount of work that a coder does now, will we need 1 tenth of the coders? Or will we simply create 10 times as much stuff? My guess is that we'll create 10 times as much stuff. The human appetite for new experiences, new entertainment, new content is basically unlimited. Our desire to have it personalized to us is basically unlimited. And I think more in general is going to replace less. But even if that correct, it will not be a painless transition. Part of what makes AI so dramatic even relative to other technologies is the speed at which the change is happening. This is not something that gets adopted

a little bit. This is something that when an individual adopts one of these new workflows and one of these new tools, they will simply never go back to the way that they used to do it. That's like nothing we've ever seen before, and it's creating an incredible urgency around this transition. In any case, I'm glad to see this type of discourse, one that doesn't seek to ban AI or lionize, it, but instead says how can we make it work for people? Start to get a little bit louder in the discussion. I hope that's something that we can contribute to, both at the AI breakdown and, of course, with Super Intelligent, whose entire mission is giving people more agency in this transformation. And I think there's a lot more people thinking like that now. But that is going to do it for

today's AI breakdown. Hope you're having a great weekend wherever you are. And until next time, peace.