The AI Daily Brief: Artificial Intelligence News and Analysis - OpenAI's COO Says AI for Business is Overhyped. Is He Right?

Episode Date: December 7, 2023

OpenAI COO Brad Lightcap recently got a big conversation going when he suggested that AI for business was overhyped. NLW explores whether he's right. Before that, Elon Musk is raising up to $1B for xAI. Today's Sponsors: Listen to the chart-topping podcast 'web3 with a16z crypto' wherever you get your podcasts or here: https://link.chtbl.com/xz5kFVEK?sid=AIBreakdown  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 discussing comments from OpenAI's COO that AI for business is overhyped. Before that on the brief, XAI is in the process of raising up to a billion dollars. The AI breakdown is a daily podcast and video about the most important news and discussions in AI. Go to Breakdown.net network for more information about our YouTube, or Discord, and our newsletter. Welcome back to the AI breakdown brief, all the AI headline news you need in around five minutes. In the wake of OpenAI's board and leadership controversy, there is certainly more openness than even there was before to alternatives to ChatGPT and OpenAI, and many players are looking to capitalize on exactly that. Now, Elon Musk did not need any sort of Sam Altman controversy

to want to wade into this space, but it does appear that he is ramping up efforts even further with his AI startup, XAI. According to new SEC filings, XAI is planning to raise up to $1 billion and has raised 135 million of that already. The first sale occurred on November 29 and involved four investors for that 135 million. Now, adding some complication to this story, earlier this morning, Elon denied these reports saying that they were not raising money for XAI right now. But again, clearly there is some amount of it going on given that there was an SEC filing around this.

What is clear is that Elon and the team at XAI continue to be extremely excited about GROC. They continue to position it as an alternative to chat GPT that has more personality, more voice, more willingness to wade into areas that are not PC. Still, when it comes to the practical benefits that make it a different offering than Bard, Claude, or Chat, GPT, it seems much more likely that the value proposition is going to be about, one, the access to Twitter slash X's incredible trove of data, and two, the fact that it will be integrated directly into that platform, starting with X premium plus subscribers. I think all on its own, having access to real-time information and conversations from X is going to mean that it has a place in this complicated tapestry, and

in many ways has a more differentiated starting point than some of the other tools. Never a dull moment around Elon, that's for sure. Now, one last little sub-story that some people are keeping an eye on. Nathan Lanz from lore.com posted about a week ago, prediction. Ilya will leave open AI and join XAI. Why? As of today, Ilya is no longer on the board of open AI after playing a role in the firing of Sam Altman. It seems likely that Ilya will either have a reduced role or be pushed out. Elon Musk has talked highly about Ilya and how he helped bring on Ilya. He has stated that this was the most pivotal hire in OpenAI's history. Also in the most recent interview with Elon at Deal Book, he stated he has tried to reach out to Ilya. I'm trying to imagine how

Ilya will be working at OpenAI with how awkward it must be that he is no longer one of the leaders, and since he seemed to have a pivotal role in Sam Altman's firing. Now, we haven't seen any sort of confirmation or even rumor of this outside speculation, but it was notable last week when Sam Altman returned to the company that he made it clear in that initial memo that Ilya's role had not yet been sorted out. Now, speaking of companies taking on the 800-pound gorilla in the AI space, AMD is trying to dethrone or at least make progress against InVideo with the launch of its new MI-300 chip. This chip has been a long time in the coming. It was announced months ago. The company AMD has said that it will contribute $2 billion to the bottom line next year, but later today

we will get a full launch event that could include appearances from companies including Microsoft, and from there we'll actually be able to see how this thing works in the real world. Now, of course, invidia is not sitting still either. They recently announced their latest chip, the H-200, which is, of course, explicitly designed for AI, although they've had to delay the release of that chip to next year. Interestingly, it doesn't seem to be that domestic competitors. Instead, it's the Chinese manufacturer or Huawei that Nvidia's CEO Jensen Huang is recently identified as a, quote, formidable AI chipmaking rival. Now, of course, part of the reason for that competition is the larger geopolitical competition between the U.S. and China, which has economic and technological dimensions,

of which AI is certainly a part, but when it comes to Washington, D.C., Axios reports that increasingly the conversations in D.C. aren't just about, A, geopolitical competition with China, or B, existential risk, but instead looking enthusiastically about the possibility that AI could increase bureaucratic efficiency. Writes Axios, D.C., a year after ChatchipT burst into the world, a once-wary Beltway establishment sees it as a tool for real government efficiency. These are dreams made of, say, faster passport renewals or AI ferreting out tax cheats and of contractors cashing in on a new Washington boom. The talk of the town has flipped from fears of the singularity to an embrace of chatbots. Blanket bans are out. Instead, President Biden recently ordered agencies to hire chief AI officers

and provide employees, quote, access to secure and reliable generative AI. Use cases abound. Northern Virginia's tech corridor is building tools to process government paperwork faster than pencil pushers. Spy chiefs want AI to one day help space satellites repair themselves. contractors are raring to harness the Uber computing power for new surveillance tools. Basically, there is now a shift to capitalize on all of this, and that is leading to lots and lots of excitement, particularly among the procurement class, where I expect a bonanza in the years to come. Now, one controversial question when it comes to AI is to what extent society will be comfortable

recreating the voices or likenesses of people who have passed on. There was a big conversation around this with the recent last Beatles song, and of course the band and the companies around them went to pains to make it clear that this was not an AI-generated John Lennon, but was an AI-restored version of John Lennon's actual vocals. Of course, this question of likenesses has also been at the heart of the sag strike, but in some ways it's hard not to feel like there is going to be an inevitable and inexorable push to create AI content using versions of the artists that we loved and lost. One new example is a sleep story from the appcom that uses an AI-generated version of Jimmy Stewart's voice, best known for his

appearance in It's a Wonderful Life for a story that they're calling It's a Wonderful Sleep Story. Now, Jimmy Stewart died 26 years ago in 1997. So a couple things to note about this particular process. The voice was recreated in collaboration with an AI voice cloning startup called Reespeacher. Basically, they employ a voice actor who performs the lines mimicking Jimmy Stewart's speech style and then overlays the AI version of Jimmy Stewart's voice on top of that. Said Alex Sertiak, the Respeecher CEO and co-founder, James Stewart is one of the most remarkable actors in U.S. history. Recreating his voice with AI was both a huge responsibility and an honor for re-speecher. It was a way to pay tribute to his incredible career and all the good memories he left to people.

While not everyone is thrilled about the use of AI in this way, Stuart's family gave their approval for this, said Kelly Stewart Harcourt, one of Stewart's daughters in a statement, we're excited for our dad to be the voice of Com's latest sleep story. It's amazing what technology can do and wonderful to see dad's legacy live on this holiday season in new ways, like helping people find restful sleep and sweet dreams. Like I said, I am quite sure that there is going to be more, not less of this type of controversy in the months to come. Now, lastly, I wanted to call out a new tool which has just launched. It's on product hunt right now. The tool is called Visual Electric, and they call it the first image generator design for creatives.

Now, what's interesting about this is that they've actually gone away from the chat-based interface, and instead created a tool that's custom purpose for a creator or designer use case. Now, what I think is interesting about this is that my guess is that it's on the leading edge of a trend that we're going to see a lot more of. which is that even though natural language inputs have become such a new force in the way that we interact with computers, it's highly unlikely that they're going to be the be-all-end-all for all use cases. As we move deeper into the phase of integration and more practical utility for these tools, you're likely to see a lot more tools like Visual Electric, where the underlying powers of generative AI, in this case image generation,

are repurposed in specific user experiences that are designed and optimized for a particular type of use case. In this particular use case, Visual Electric is arguing that the chat interface on its own is not necessarily what designers really need. So far, the excitement and buzz around the company suggests that they might be on to something, and I'm looking forward to playing around with it a little bit later today. However, for now, that is going to do it for today's AI breakdown brief. Next up, the main AI breakdown. Quickly a brief word from today's sponsor.

As a listener of this show, I suspect you like to stay up to date on all things AI and tech, which is why you have to check out the chart-topping podcast Web3 with A16Z Crypto. Produced by venture firm Andresen Horowitz, Web3 with A16Z is the perfect companion podcast to the AI breakdown. Web3 with A16Z crypto is your definitive resource for the future of the internet, whether you're interested in the convergence of AI and crypto or simply curious about what's next. If you need a place to start, they recently released an excellent episode with Stanford Cryptography Professor Dan Bonay and former Google X engineer Aliya in conversation with host Sonal Choxi about the intersection of AI and Crypto.

From fighting deepfakes and proving humanity to large language models like ChatGBT, they cover it all. I highly recommend checking it out, especially if you'd like to learn more about how AI and crypto will impact our everyday lives. Beyond Crypto and AI, this show is for creators seeking more ways to truly own their work, for business leaders trying to prepare for the future today, and for innovators exploring trending tech topics. Don't miss out. Follow Web3 with A16Z Crypto on Apple Podcasts. Spotify, or your favorite listening app. Welcome back to the AI breakdown.

Today we are discussing a theme which I've seen come up a little bit and specifically in the context of recent statements from OpenAI COO Brad Lightcap. The gist of it was summed up in a verge headline which read, OpenAI CO thinks AI for business is overhyped. AI can't transform your business overnight, he said. Now, of course, the denizens of X slash Twitter, especially those who have held this type of opinion before,

grabbed onto this statement and used it to confirm what they had been saying all along. See, for example, Gary Marcus, who said, Nice to see I'm not the only one thinking AI is overhyped. So what I want to do today is use this as a jumping off point for a larger conversation. To do so, however, let's start with what Brad Lightcap actually said. The context for this was an interview with CNBC. And basically, what Lightcap was saying is that the companies that were coming to OpenAI to explore Enterprise AI solutions didn't always have an accurate or realistic sense of what

AI could actually do for them. Writes the Verge, LightCAP said companies have approached Open AI expecting generative AI to solve many problems, dramatically cut costs, and bring back growth that they're struggling. He said that while AI could improve more, quote, there's never one thing you can do with AI

that solves that problem in full, and that the technology is still in its infancy. He added that AI is still in the experimental phase where it is yet to become part of critical tools and applications. So, this is actually slightly more nuanced, surprise, surprise, than the headline would make it seem. Like, HAP isn't really actually saying that AI is overhyped, but that enterprise or business users of AI who don't really understand the technology yet have mistaken expectations of what it can

actually do, or what it can do on its own, or what it can do without having to be integrated into workflows. This may seem like semantics, but I don't think it is. If the question is, are the AI tools that exist right now, all on their own, when dumped into a business context, without any real consideration, just going to solve all of the problems of business? The answer is obviously no. But no hype is actually claiming that that's true.

What the hype is claiming is that these tools are going to have an absolutely transformative effect on a huge array of functions inside business, change processes in radical ways, lead to new efficiencies, lead to redundancies of skill sets, and ultimately represent a total phase shift in how businesses function. The answer to that question is, I believe, unequivocally yes. And frankly, the assumption that that would happen overnight, or without real human intention, in other words, just passively, isn't really an opinion that I think a lot of people have. So let's move on from the question of overhyped or not, and actually get into how businesses

and employees are thinking about what this change is going to look like. So let's look at some recent studies that get at attitudes around AI. One comes from EdX back in October and was a survey of both CEOs and employees in America asking about their opinions on AI and how it would change the workforce. Now, the big banner headline that captured a lot of people's attention was that the CEOs who were surveyed estimated that 49% of the skills that exist in the workplace today will not be relevant within two years by 2025. 47% said that their workforces were unprepared for the future workplace. So what does that change actually look like? Well, executives in that survey estimate that within the next

five years, their organizations will eliminate 56% of entry-level knowledge worker roles, and 79% of executives predicted that entry-level knowledge worker jobs simply won't exist anymore. And just to be clear, that this isn't just for low-level workers, those same executives also said that 56% of their own roles will be completely or partially replaced by AI. 47% of C-suite executives believed that most or all of the CEO role should be automated or replaced by AI, including 49% of CEOs agreeing with that as well. Overall, 92% of the executives felt that it was important to improve their AI skills within the next one to two years, and 79% of executives said that they fear that if they don't learn how to use AI, they will be unprepared for the future of work. Now, what about

regular people. In another even more recent survey from Spokio, exactly two-thirds, 66.6% of the workers surveyed, said that AI could carry out most of their workplace functions. Now, this is generating two very different types of responses in people as well. Just about 75% said that they were concerned about AI's impact on their industry overall, but 78.1% also said that they believe that AI could reduce some stress on the job. 76.7% said that they think that AI will eventually reduce the number of working days in the week for the average American. They also think that employers should be involved in re-skilling their teams. 79.1% said that they think employers should offer training for chat CBT and other tools and other tools like it. And so again, overall,

you're getting a picture where professionals all the way up to the C-suite have this sense that AI is going to have dramatic impacts on what they do. Now, there are, of course, two possibilities. The first is that they're just responding to media headlines about this changing everything. It's not hard to find this sort of headline. For example, from November 16th in Fortune magazine, Altimiter CEO Brad Gersner warns AI will cause, quote, the largest displacement of human labor in the history of capitalism. So are these responses that are showing up on surveys, just artifacts or reaction to that sort of bombastic and scary headline? It's totally possible, and it's hard to think that that wouldn't have some impact overall on the sense of the magnitude

of this disruption. I would argue, however, that these are likely not just people who are reading articles and responding in kind. I don't think it takes too many interactions with chat chisb-t or mid-jorney or any one of these other tools, which the first time you use them feel like literal, actual magic, to have a sense that how you work and probably what you work on are forever going to be shifted. Now, what that doesn't mean is that that's going to happen tomorrow. There are huge numbers of reasons why this change will be perhaps slower than we think, standing in the eye of the storm as we are. Some of those are the natural inertia of business and technology. Some of that is compliance, where it's likely that humans will be required to be involved in things for a lot longer than they

might need to be technically speaking. Whatever the case, in terms of the way that AI integrates into business, it's likely to take longer than the speed at which it changes from a technology capability perspective. It's also not going to happen passively. We're at a phase now, where the way that AI is starting to impact business is not necessarily company-wide strategies or formally updated workflows and processes, but instead individual employees adopting these tools and using it to enhance their own productivity or increase their capabilities or expand the range of what they can do, and slowly bringing that into the broader company such that some of those best practices can become normalized and regularized. In other words, we're in a highly organic phase even as these

big companies all scramble around in boardrooms and in conversations with people like OpenAI's Brad Lightcap, trying to figure out what their AI stack and AI solutions should look like. I think Professor Ethan Mollock had a really insightful tweet on this. On November 16th, he wrote, We're going to go through a fairly extended period where work has changed dramatically for many people thanks to AI, but companies will not notice or acknowledge it. This is exactly what is happening in schools.

AI broke homework. Most instructors haven't changed anything they do. I think this captures the essence of it. I don't think that you can interact with chat GPT and then as any sort of knowledge worker think that it's not going to impact what you do and how you do it. But there will be a period of time where the counterpoint is that it doesn't feel like all that much has changed. So when it comes back to this question of whether AI for business is overhyped,

my answer very clearly is that no, it is not. It's just that, one, we don't have the precise time scales. And two, businesses are still going to have to be actively involved in their own disruption. AI won't just happen to them. People inside companies will bring it in, start to figure out how to integrate it in ways that actually drive value, be that efficiency gains or productivity gains, or actual real bottom line value, or, alternatively, upstart competitors who are using that technology, will come in and challenge the incumbents in ways that forces them to change or go away.

It's going to be a big, long, messy process of creative disruption, but like it or not, it's happening. Anyways, guys, that will do it for today's AI breakdown. Appreciate your listening or watching as always. Until next time, peace.