Tech Brew Ride Home - Thu. 02/08 – Google Releases Gemini Ultra 1.0

Episode Date: February 8, 2024

Google has released Gemini Ultra 1.0, renamed Bard as Gemini and looks to be replacing Google Assistant with Gemini. Disney has invested in a big stake of Epic Games to get at Fortnite IP. Leaked images of the Pixel Fold 2. And what if OpenAI is facing the same strategic dilemma that Mark Zuckerberg was never able to overcome? Sponsors: Robinhood.com/boost Links: Google’s AI now goes by a new name: Gemini (The Verge) Google Assistant Just Got Supercharged With AI. It Might Be the Biggest Update in Google's History. (Gizmodo) Google Prepares for a Future Where Search Isn’t King (Wired) Google Joins Effort to Help Spot Content Made With A.I. (NYTimes) Disney to take $1.5 billion stake in Epic Games, work with Fortnite maker on new content (CNBC) China had "persistent" access to U.S. critical infrastructure (Axios) Exclusive: This could be the Google Pixel Fold 2 (Android Authority) OpenAI Shifts AI Battleground to Software That Operates Devices, Automates Tasks (The Information) Learn more about your ad choices. Visit megaphone.fm/adchoices

It's peak pollination season, and my business is scaling fast. To keep the nectar flowing, I need a phone plan with top priority data speed. That's why I chose GoogleFi Wireless. My connections stay strong even when the hive is buzzing. Plus, unlimited plans start at $35 a month. Now, that's a deal that doesn't stay. Explore GoogleFi Wireless plans today. Plus taxes and government fees.

GoogleFi Wireless is not subject to data traffic deprioritization during times of high network usage. Welcome to the TechMeme Right Home for Thursday, February 8th, 2024. I'm Brian McCullough today. Google has released Gemini Ultra 1.0, renamed BARD as Gemini, and looks to be replacing Google Assistant with Gemini eventually. Disney has invested in a big stake in Epic Games to get at Fortnite IP, leaked images of the Pixel Fold 2, and what if Open AI is facing the same strategic dilemma that Mark Zuckerberg was never able to overcome?

Here's what you miss today in the world of Tech. Google has released Gemini Ultra 1.0, the largest and most capable version of its LLMs, and has renamed Bard 2 Gemini, which now has a dedicated Android app, quoting the verge. Gemini's mobile apps will likely be the place most people encounter the new tool. If you download the new app on Android, it can set Gemini as your default assistant, meaning it replaces Google Assistant as the thing that responds when you say, hey, Google, or long press your home button. So far, it doesn't seem Google is getting rid of assistant entirely, but the company has been deprioritizing assistant for a while now,

and it clearly believes Gemini is the future. I think it's a super important first step towards building a true AI assistant, says Sisi Sao, who runs Bard, now Gemini at Google. One that is conversational, it's multimodal, and it's more helpful than ever before, end quote. There's no dedicated Gemini app for iOS, and you can't set a non-Sy assistant as the default anyway, but you'll be able to access all the AI features in the Google app. And just to give you a sense of how important Gemini is to Google, there's going to be a toggle at the top of the app that lets you switch from Search to Gemini. For the entirety of Google's existence, search has been the most important product by a mile. It's beginning to signal that Gemini might matter just as much. For now, by the way,

Google's in Search AI is still called Search Generative Experience, but it's probably safe to bet that'll be Gemini eventually too, end quote. On Gemini Ultra, quoting Gism Moto. According to Google, Gemini Ultra is the most advanced AI on the market. The company says Gemini Ultra is the first AI model to outperform human experts on a standardized test called MMLU, massive multitask language understanding, which measures in AI's knowledge and problem-solving capabilities and a combination of 57 subjects such as math, physics, history, law, medicine, and ethics. Google's new AI business is shaped a lot like ChatGTPT, the free version of Gemini runs on the basic Gemini Pro model, just like the free Chat-Cepti-T-T-T-tier run. Just like the free Chat-Cepti

tier runs on GPD 3.5. If you want the full capabilities of Gemini Ultra, it costs $19.99 a month a penny shy of what OpenAI charges for GBT4. Gemini Ultra comes with other perks as well. It's now rolled up into a new premium tier of Google One, the subscription service that gives you more storage and other perks. Gemini Ultra comes as part of the new Google One AI premium plan, which includes all the perks of the 2 terabyte storage plan. You can try a free two-month trial if you want a preview. If you don't want AI, the regular 2-Tabite plan still costs $9.99 a month, end quote. As Gizmodo points out, with Gemini on your phone, you're now carrying around basically a full-fledged AI device. But to be clear about that rebranding, the $20 per month AI premium plan is on top of the

existing 2-terabyte premium Google One Plan, which was $999, but now it offers Gemini advanced with Ultra 1.0. And to be clear, they've also fully rebranded Duet AI as Gemini for Workspace. Google is all in on Gemini going forward, not duet, not barred Gemini. As for their business model going forward, quoting a wired interview with CEO Sundar Pichai, Pichai says that Google is focused right now on getting the generative AI experience right, but that he is, quote, open to possibilities around both paid and ad-supported generative AI experiences. He declines to say whether the paid Gemini offering will remain totally ad-free, but pointed to another Google-owned product where it's possible to banish ads entirely. Quote,

YouTube has been a very good example of this, Pichai says. A reference to the paid ad-free tier that YouTube started experimenting with several years ago, ads allows us to give products to more people, but there will be cases of subscriptions that allow people to get a different experience, he adds. I can imagine the same user going back and forth between free search and a Gemini subscription.

In other words, generative search would no longer be a side dish to search, but a main menu item, albeit a more expensive one, end quote. Real quick, Google also announced this morning that it has joined the C2PA steering committee to develop a standard to label AI content via metadata alongside Adobe, the BBC, Microsoft, Sony, and others. This is the C2PA that I told you about meta joining recently, quoting the New York Times. The tech giant said on Thursday that it was joining an effort to develop credentials for digital content, a sort of nutrition label that identifies when and how a photograph, a video, and audio clip,

or another file was produced or altered, including with AI. OpenAI said this week that its AI image generation tools would soon add watermarks to images according to the C2PA standards. Beginning on Monday, the company said images generated by its online chat bot, chat GPT, and the standalone image generation technology, Dali, will include both a visual watermark and hidden metadata designed to identify them as created by artificial intelligence. The move, however, is not a silver bullet to address issues of provenance. OpenAI said, adding that the tags, quote, can easily be removed either accidentally or intentionally, end quote. In 2024, as billions of people prepare to vote in significant global elections, identifying the source of online content and tracking its modifications, is becoming a major concern for policymakers and digital monitors. The C2PA initiative has been embraced by a wide range of stakeholders, including news outlets like the New York Times and even camera manufacturers. financial institutions, and advertising firms. Guess what? You might see some Fortnite characters the next time you go to Disney World.

That's because Disney plans to invest $1.5 billion for a stake in Epic games and collaborate on a new, quote, universe connected to Fortnite with content from Disney, Pixar, and more. Quoting CNBC, Disney did not say what the valuation of Epic, a private company, would be after the media company's funding. In an interview with CNBC's Julia Bort's, Disney CEO Bob Eiger called the investment, quote, probably our biggest foray into the game space ever, which I think is not only timely, but an important step when you look at the demographic trends

and where Gen Alpha and Gen Z and even millennials are spending their time in media, he said. The partnership comes after Disney had success licensing figures such as Spider-Man for Blockbuster video games and collaborated with Epic to bring characters from Marvel, Star Wars, The Nightmare Before Christmas, Tron, and more to Fortnite. The deal also extends a string of major partnerships for Epic, Fortnite has recently collaborated with Lego for a survival crafting game within the gaming platform similar to Minecraft. It also launched Fortnite Festival, a rhythm game from Harmonics, which created the game rock band. Disney was one of the first companies to believe in the potential

of bringing their worlds together with ours in Fortnite, and they use Unreal Engine across their portfolio, said Epic Games founder and CEO Tim Sweeney in a statement. Now we're collaborating on something entirely new to build a persistent, open, and interoperable ecosystem that will bring together the Disney and Fortnite communities, end quote. What is somewhat interesting to me is nowhere in any of those statements, does anyone mention the term metaverse, but that's exactly what they're talking about, right? The so-called Five Eyes Intelligence Alliance, which is the U.S., UK, Australia, Canada, and New Zealand, have issued an advisory saying that China-backed hacking group Volt Typhoon has had access to some major U.S. infrastructure for over five, five,

years. Quoting Axios. The hacking campaign laid out in the report marks a sharp escalation in China's willingness to seize U.S. infrastructure going beyond the typical effort to steal state secrets. According to the advisory, China-backed hacking group Volt Typhoon has been exploiting vulnerabilities in routers, firewalls, and VPNs to target water, transportation, energy, and communication systems across the country. The group has relied heavily on stolen administrator credentials to maintain access to the systems, and in some cases it has maintained access for, quote, at least five years per the advisory. Volt Typhoon has been seen controlling some victims' surveillance camera systems, and its access could have allowed the group to disrupt critical energy and water controls.

Of note, Volt Typhoon uses so-called living off the land techniques that limit any trace of their activities on a network, making the actors more difficult to detect. U.S. officials are increasingly worried China will launch destructive cyber attacks either during or in the lead-up to a possible Chinese invasion of Taiwan. Authorities in Canada, Australia, and New Zealand contributed to today's advisory, citing concerns that China is also targeting organizations in their countries. Intelligence officials have been ringing alarm bells about Volt Typhoon for nearly a year. Last May, Microsoft and the U.S. government warned that Volt Typhoon had been positioning itself to launch attacks on infrastructure across the country, including water utilities and ports.

This month, officials said they had successfully thwarted Volt Typhoons access to these networks, but warned that the group had shown a willingness to keep looking for new ways in, end quote. Once again, over in Android land, stuff just leaks. A new image has surfaced allegedly showing the Google Pixel Fold 2, including a narrower cover screen, a squarer inner screen, aspect ratio, and redesigned camera bump. Quoting Android Authority. Although we were only sent a single, blurry, heavily censored photo of the Pixel Fold 2, we can tell right away that Google's next foldable phone has a narrower,

cover screen. We don't have the exact dimensions, but comparing this photo with a photo we took of the first-gen pixel fold at a similar angle makes the difference quite clear. Our source told us that the cover screen is narrower, but more importantly, the inner screen's aspect ratio is closer to a square. That suggests the fold two's overall form factor might be similar to the one-plus open, which has screens that are in between the pixel fold and the Samsung Galaxy Z-Fold 5 in terms of size. The narrower form factor of the pixel fold two may come as a surprise to some, but there's a reason why most other foldable phones aren't as wide as Google's first-gen foldable. A lot of Android apps still aren't optimized for tablets. By making the first-gen-pixel fold as wide as it

is, Google had to employ software workarounds to deal with apps that refuse to run in landscape mode. The company even developed an entirely new OS feature in Android 14 QPR1 that lets users force apps like Instagram to go full screen. Had Google chosen to make the first-gen-gen-pixel fold narrower, then it wouldn't need to resort to such tricks, as a slimmer form factor would cause most apps to show their regular phone optimized UI instead of their tablet optimized UI if one even exists. The new form factor isn't the only change visible in this leaked image. The photo also reveals that the camera bump on the rear could be getting a redesign. The pixel fold two appears to have an isolated camera island on the top left consisting of four sensors that likely

include a main wide angle lens, a secondary ultra-wide angle lens, a tertiary, periscopic telephoto lens, and an unknown cordonary lens, end quote. Just yet. Just yet. Yesterday, another leaker told Android Authority that the PixelFold 2 might shift its chipset to the TensorFlow G4 instead of G3. The rumors also suggest that the PixelFold 2 could get a bump up to 16 gigabytes of RAM. Finally, today, some whispers about OpenAI's possible evolving strategy. Sources are telling the information that OpenAI is developing two types of agent software, one to automate tasks by effectively taking over a user's device,

and the other for web-based tasks, taking over devices. Now you can see why Sam Altman might want to make his own hardware. Quote, OpenAI is developing a form of agent's software to automate complex tasks by effectively taking over a customer's device. The customer could then ask the chat GPT agent to transfer data from a document to a spreadsheet for analysis, for instance, or to automatically fill out expense reports and enter them into accounting software. Those kinds of requests would trigger the agent to perform the clicks, cursor movements,

text typing and other actions humans take as they work with different apps, according to a person with knowledge of the effort. OpenAI will have to temper fears about agents that access users' computers, given that such a takeover might remind some people of malware, which is known to elicitly seize control of computers and steal their data. Open AI, a leader in generative AI that investors recently valued at $86 billion, is developing another class of AI agent that would handle web-based tasks, such as gathering public data about a set of companies, creating itineraries under a certain budget or booking flight tickets, said a person with knowledge of that effort. Google and Meta have said they are developing similar types of agents, powered by conversational

AI known as large language models. OpenAI's plans for the agent that takes over people's computers will require the user's permission to work. To operate in a personalized fashion and respond quickly the way Apple Siri does on the iPhone, the prospective open AI computer using agent may need to be partly stored on users' devices. The company may also need to get permission from users to train the software on personal data, such as an individual's emails and contacts, as well as information stored in business apps like Word and Google Docs. In contrast, people today access chat GPT through a website or mobile app, and all of its computations happen in the cloud, specifically through Microsoft's Azure servers.

It isn't clear when OpenAI plans to release its agent products, which have been in development for more than a year, but some of its employees have hinted at their importance. Last month, Ben Newhouse, an OpenAI employee who has worked on computer-using agents at the startup, according to a person familiar with his role, posted on X that he was hiring for his team and, quote, building what I think could be an industry defining zero-to-one product that leverages the latest and greatest from our upcoming models. He didn't elaborate. Peter Weelander, Open AIs, vice president of product, remarked on X that the product Newhouse described, quote, will change everything, end quote. So they're talking about computers there, like

desktops and laptops. But again, let's extrapolate a bit into the future. Let's imagine your Open AI. As you're looking down the road, are you seeing you might run into the same brick wall Mark Zuckerberg and Meta did all those years ago? You see AI on device, not in the cloud as the future of computing, but you don't own a mobile platform. You don't own the hardware. You don't make or control the ecosystem that would allow you to do on-device AI computing. Google and Apple do, and it would be far, far from certain that they would want you to do things like take over their user's devices. Crazy strategic problem, don't you think? Nothing for you today. Talk to you tomorrow.