The AI Assistant and Agent Battle Heats Up

Today on the AI Breakdown, we're looking at a big day in AI agents and assistants and how the competition is shaking out for the future. 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, One quick note over here on the east coast we have been dealing with an ice storm for the last couple days and and my Internet has been barely working. So much so that I've spent the last six hours or so offline and was only able to get a main part of the episode done. So apologies that this episode does not have a brief. We'll be back hopefully fingers crossed to our normal format tomorrow. Welcome back to the AI Breakdown. Today we are talking about a theme that has promised to be a big one for 2024 and is coming out of the gate in an even more significant way than we thought. That is the convergence of AI assistants. In other words, incredibly powerful personal assistant applications that can actually go out and do tasks on one's behalf and AI hardware yesterday saw the launch of the Rabbit R1. The company had been hyping the launch for the past several weeks and unveiled yesterday what people immediately recognized as a very Jobsian style keyNote presentation, about 20 minutes long, explaining exactly what the Rabbit R1 was all about. Now there are actually two parts of Rabbit's presentation. The first was what lies underneath, which they call Rabbit os, which is underpinned by something they're calling a large action model. They write the Large Action model is a new type of foundation model that understands human intentions on computers with lam Rabbit os, understands what you say and gets things done. And so basically a big part of their presentation was trying to draw a distinction between large language models and large action models, pointing out, for example, that while ChatGPT is very good at finding information, it's not set up currently really to take actions on a user's behalf. The second thing that Rabbit revealed, however, was the R1, a totally new type of device that does away with many of the conventions of personal computing and is another contender for what a next generation AI native mobile device is going to look like. So the R1 has a few different features. The main way that you interact with the R1 is by talking. There's a single push to talk button which as soon as you press it you're off to the races. There's a 360 degree rotational eye which gives the device the ability to see and interact with the world visually there's an analog scroll wheel and a USB C port and a SIM card slot. Now a lot of the way that the R1 will work is that you'll use a web based portal to connect various services that you use. Think for example Spotify to the R1 so that this device itself becomes kind of like a voice activated control center for all of the applications that you might otherwise use. Now inherent in the presentation was a critique of the existing form factor, the way that applications are organized currently and have been for the last 15 years of smartphones. Here's how the CEO described it.

Like iPhone and Android phones. These guys been here for years and we've grown tired of them. The problem with these devices however, is not the hardware phone factor, it's what's inside the app based operating system. Each time you want to do something, you fumble through multiple pages and folders to find the app you want to use and there are always endless buttons that you need to click. The smartphone was supposed to be intuitive, but with hundreds of apps on your phone today that don't work together, it no longer is. If you look at the top ranking apps on app stores today, you'll find that most of them focus on entertainment. Our smartphones has become the best device to kill time instead of saving them. Recent achievements in large language models, however, or LLMs, a type of AI technology, have made it much easier for machines to understand you. The popularity of LLM chatbots over the past years has shown that the natural language based experience is the path forward.

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 a 16z crypto. Produced by venture firm Andreessen Horowitz, Web3 with a 16z is the perfect companion podcast to the AI breakdown Web3 with a 16z 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 Bonet and former Google X engineer Aliyah in conversation with host Sonal Chokshi about the intersection of AI and crypto. From fighting deepfakes and proving humanity to large language models like ChatGPT, 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 a 16z crypto on Apple podcasts, Spotify, or your favorite listening app. Now one of the things that I was most impressed with when it came to their approach was that they sort of put their money where their mouth is when it came to the critique of how existing mobile services work. So whereas for example, the Humane PIN is going to be something like $699 plus a cell phone plan on top of that, the rabbit is $199 for the device with no subscription required. Now when it came to the community's reaction, there were a couple different categories. Some thought their presentation was great. Jason Levin said, I've done marketing for 30 plus startups. Ra Rabbit's launch of their AI hardware was genius. To be honest, the founder might be the next Steve Jobs. Here's the hidden marketing tricks you definitely didn't notice. He goes on to point out the critique of the endless buttons on smartphones as well as the critique of LLMs understanding language but not being able to take actions. He also pointed out that it was a super savvy move to collaborate with what he called Silicon Valley Nerd's favorite product designers. Teenage Engineering. Teenage Engineering are better known for their pocket and portable synthesizers. They're totally reimagining the music creation process and they really do have kind of a cult following. Now. The biggest category of critique that I saw is why couldn't this just be an iPhone app? Sergey Galgan put a little Rabbit icon at the bottom of an iPhone and said, is there a universe in which I have this and don't have to spend $199 to carry around another device in my pocket? Gabe Schnitzel created a little app icon and said, I just saved you guys a bunch of time. Plastic and spinny wheels. And this is just the tip of the iceberg of that particular critique. Matthew Berman, however, offered a defense which again echoed a lot of themes that I saw. He wrote why I'm excited about Rabbit R1 new device. I'm a techie, so I love it. Natural language based interface removes complexity in the form of apps. Rabbit R1 offers a preview of the future, one where apps are gone and operating systems are abstracted away by LLMs. Could this have just been an app today? Maybe. If so, it would have to navigate current app Store restrictions Apple will likely launch something like this built into iPhone. It'll be good, but I enjoy trying new form factors and the price point is fantastic and I love the design by teenage engineering. Bravo to a company willing to try something completely new. I think Matthew's point that this offers a preview of the future in which apps are gone and operating systems are abstracted away by LLMs is an extremely salient one. It may be that there isn't room in the mainstream for yet another category of consumer device that sits alongside a phone, but it also might be that the design experiments that show up in things like the Rabbit ends up shaping what the iPhone 20 looks like a few years down the line. Now, like I said, what made yesterday interesting was that the news around AI assistance and AI hardware wasn't just from Rabbit. First of all, Amazon Alexa got a set of new generative AI powered experiences. Frankly, the updates were kind of minor. They were debuted at CES and they just showed off a set of three examples of how different AI companies like character AI are collaborating with Amazon Alexa for a new generation of skills and experiences. But to me this felt very much like prelude to much bigger changes that I think are probably coming down the pipeline for Amazon in the future. Meanwhile, Humane, which had been super buzzy and which unveiled itself last November, announced that it was laying off 4% of its employees before the release of its AI pin. A lot of the commentary was fairly incredulous, fuko Capital Guy writes. Humane already beginning layoffs before the product is even shipped. Now the founders of the company argued that this was a restructuring based on shifting phases of the company's life. But I would say that broadly the sense on Twitter was that it might be positioning for a world in which the Pin hadn't been quite as well received as they would have hoped. I think, of course, ultimately that 4% is a fairly small part of the workforce and that smart startups today are constantly in a reevaluate and revisit cash flow management kind of perspective. But the signal it sends is at least a little bit rough. Now, one of the David versus Goliath upstarts in the AI assistance slash hardware space is of course Tab. Yesterday the company announced that it had raised a $1.9 million seed round. So what does this Superpower Tamagotchi actually do? Basically, the Tab's approach to AI is to listen to everything that you say and everything that people say around you and to try to give you access to basically a second brain. It's closer to a pensieve from Harry Potter than a personal assistant. In fact, in an interview with Fast Company, Schiffman said, tab is not an assistant, period. I'm not building something that's going to connect to Notion or your emails anytime soon. I'm solely building a friend that morphs into your creative partner, life coach or therapist as needed. What I'm trying to do is create a new relationship in your life. Radical transparency without concern of judgment. I think this is a relationship people used to have with God but is lacking in the modern world. You gotta love the ambition there. And yet still there was more news in this broadly defined agent or assistant space. Multion Founder Divg Tweets we came out of stealth today very excited to officially announce ourselves Multeon our vision for AI and Agents and thank all of our many backers and investors. In their first official blog post, they write, Multeon is an AI agent that takes actions and interacts with the digital world to tackle mundane tasks that people would rather delegate to an assistant. Multiyan frees up the human race from the mundane, boring but necessary tasks that would take hours or days to accomplish so we can more fully enjoy and live our lives. Multeon will book your dream travel itinerary in minutes, or reserve a table with six of your best friends at your favorite restaurant and confirm the best date and time for everyone in just minutes. What used to take hours and even not be possible to do is now handled by Multion in a matter of minutes. In many ways, I think it's reasonable to look at Multion as a software driven approach to some of the same problems that Rabbit is trying to solve, although the very presence of hardware obviously makes it a different type of thing. Still, if 2023 was the year that language models became a mainstream force, if entrepreneurs have their say, 2024 is going to be all about AI agents and assistants. Exciting stuff to watch as we kick off this January, but for now that is going to do it. For today's AI breakdown. Until next time, peace.