Tech Brew Ride Home - Wed. 05/15 – Google’s I/O Announces.

Episode Date: May 15, 2024

All the announces from Google’s massive I/O event yesterday, including, yes, generative AI summaries are fully coming to Google Search. Plus, the camera-based AI system they teased that looks really cool. Ilya Sutskever officially leaves OpenAI. And is crypto the only place left where you can raise a billion-dollar seed round? Sponsors: Lumen.me code RIDE for $100 off Links: Google rolls out AI Overviews in US with more countries coming soon (SearchEngineLand) Can Google Give A.I. Answers Without Breaking the Web? (NYTimes) Google's Project Astra uses your phone's camera and AI to find noise makers, misplaced items and more. (Engadget) OpenAI co-founder Ilya Sutskever says he will leave the startup (CNBC) TikTok creators sue U.S. government over potential ban (Washington Post) Humanity Protocol Becomes Crypto’s New Digital Identity Unicorn (Bloomberg) Learn more about your ad choices. Visit megaphone.fm/adchoices

On April 4th, 2023, around 2 in the morning, a man was found stabbed multiple times on a sidewalk in downtown San Francisco. Hey, who did this to you? What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. Welcome to the Tech meme right home for Wednesday, May 15th, 2020. I'm Brian McCullough today.

All the announces from Google's massive I-O event yesterday, including, yes, generative AI summaries are fully coming to Google search, plus the camera-based AI system they tease that looks really cool. Ilya Sesekever officially leaves Open AI and is crypto the only place left where you can raise a billion-dollar seed round? Here's what you miss today in the world of tech. I regret not holding the show till late. later on Monday for that OpenAI event, but I'm glad I didn't hold the show yesterday for the Google I.O. keynote.

There was just so, so much as expected. In real time, I would have had zero ability to filter out what was actually important. I kind of empathize with Kyle Carr Mitchell, who said this on threads. Compared to WWDC or what we saw from OpenAI yesterday, Google I.O. continue to feel like the presentation day at the end of a hackathon. Don't get me wrong, hackathons are awesome, and those presentation days are always mind-blowing. But that's what Google is, the world's largest advertising sales company with a huge side gig of brilliant engineers doing years-long hackathons on things without anyone ever really putting together a picture of a cohesive product. It's exhausting, in aggregate, end quote. But with the benefit of 12 hours

to sort it all out, here's everything that happened along with the most important things that happened. The most important one first. Who says Google will be disrupted because the web will be obviated by others who do search by just summarizing web pages instead of pointing you to web pages. It's that head-tapping meme can't be disrupted if we obviate the web ourselves first. Google plans to roll out AI overviews in search to all US users by next week and to more countries soon. They'll be available to the one billion plus users of Google search by the end of 2024. Quoting search engine land, AI Overviews was introduced as part of the search generative experience last May. It was opt-in within the

Google Search Labs until Google began testing AI overviews in the wild for a subset of users in March 2024. AI overviews gives answers to queries using generative AI technology powered by Google Gemini. It provides a few snippets of an answer based on its understanding of queries and the content it found on the topic across the web. You won't see AI overviews for all your searches. AI overviews are reserved to answer more complex questions where, Google feels it can add value beyond the search results. If AI Overview doesn't add value to what Google search shows by default, Google will not show an AI overview.

Google would not say what percentage of queries will generate an AI overview. I asked numerous times and was told we won't share those figures. I asked if we would see AI overviews only when ads are not displayed, and I was told that was not the case. The types of queries AI overviews will show for have less to do with ads and more to do with adding value to the search results. said. End quote. More from Kevin Roos in the New York Times. AI overviews will give users concise summaries of whatever they're looking for, along with suggested follow-up questions and a list of links

they can click on to learn more. Users will still get traditional search results too, but they'll have to scroll further down the page to see them. Publishers are right to be spooked. If the AI answer engine does its job well enough, users won't need to click on any links at all. Whatever they're looking for will be sitting right there on top of their search results. And the bargain on which Google's relationship with the open web rests, you give us articles, we give you traffic, could fall apart. Google executives put a positive spin on the announcement on Tuesday saying that the new AI overviews would improve the user experience by, quote, taking the legwork out of searching. But that legwork pays for a lot of journalism and a lot of other types of online

media, fashion blogs, laptop reviews, restaurant listings, without which the internet would be far less useful. If Google's AI overviews starve these websites of traffic, what will happen to them? And if big chunks of the web were to vanish altogether, what would be left for the AI to summarize, end quote? Also, what would a podcaster have left to summarize on their daily show? We're about to run the biggest experiment on the web we've ever run, and we're running it like next week. Let's run down more of the announces from I.O. So I can feel like I told you everything. Google also announced a search filter named Web to show only text-based links in results.

So maybe this is an AI search opt-out feature, kind of. But also, it will remove forums, videos, news, images, and all that other stuff yet you get in results because they pram them in there for so long. Google also announced a private preview of a new Gemini 1.5 Pro version that can take up to 2 million tokens in at once, twice its predecessor, and rivaling the likes of Anthropics Claude 3. They announced a Gemini 1.5 Flash, which is more lightweight and cheaper than Gemini Pro, but has the same multimodal capabilities and one million token context window. Gemini Pro is coming to workspace labs, letting users summarize emails by topic and more, like Compare My Roof Repair Bids shipping later this month. Ask Photos with Gemini is a Google Photos feature that can answer questions like, show me the best photo from each national park I visited. Music AI Sandbox is a suite of music tools that include the ability to create loops via text, though they didn't give a public launch date for that.

They showed off Vio, a text-to-video model that outputs at 1080P resolution, and Imagine 3, its highest quality image generation model with improved text rendering. Circle to Search can now solve physics and math problems, and will expand later this year to diagrams and graphs powered by LearnLM models. Google said Gemini Nano is now built into Chrome on desktop with plans to use it to power features like the existing Help Me Write tool, Gemini Advanced Subscribers, will soon be able to make personalized Gemini chatbots called Gems with distinct personalities and abilities. Google also unveiled additions to Gemma, its family of open models, including Gemma 2, a 27 billion parameter model, debuting in June, and PolyGemma, a vision language model. They open source the code for

Project GameFace, their hands-free gaming mouse, controlled by head movement and facial gestures. They open-source that to Android developers. There's now AI-powered scam call detection on Android, a chatty new voice mode with its Gemini Live feature, and voice and video search with Google Lens. Then, the thing that has been passed around a lot online, they also teased Project Astra, a camera-based AI app to process visual data in real time. They have a video showing a person using their phones camera to show the AI a bunch of things, which it not only identifies, but interrogates. That video finished with that same person, putting on a pair of smart glasses and then querying the feature suggesting Google Glass could make a comeback.

But even if not, seriously, smart glasses as a key input feature for AI and the future of wearables in general is really coming into focus, I think. Quoting in Gadget. According to a video that Google showed during a media briefing yesterday, Project Astra appeared to be an app which has a viewfinder as its main interface. A person holding up a phone pointed its camera at various parts of an office and verbally said, tell me when you see something that makes sound. When a speaker next to a monitor came into view, Gemini responded, I see a speaker, which makes a sound. The person behind the phone stopped and drew an on-screen arrow to the top circle on the speaker and said, what is that part of the speaker called?

Gemini promptly responded, that is the tweeter. It produces high-frequency sounds. Then, in the video that Google said was recorded in a single take, the tester moved over to a cup of crayons further down the table and asked, Give me a creative alliteration about these, to which Gemini said, Creative crayons, color cheerfully. They certainly craft colorful creations. The rest of the video goes on to show Gemini and Project Astra identifying and explaining parts of code on a monitor, telling the user what neighborhood they were in based on the view out the window.

Most impressively, Astra was able to answer, Do you remember where you saw my glasses, even though said glasses were completely out of frame and were not previously pointed out? Yes, I do, Gemini said, adding your glasses. were on a desk near a red apple. After Astra located those glasses, the tester put them on, and the video shifted to the perspective of what you'd see on the wearable. Using a camera on board, the glasses scanned the wearer's surroundings to see things like a diagram on a whiteboard. The person in the video then asked, What can I add here to make this system faster? As they spoke, an on-screen waveform moved to indicate it was listening, and as it responded,

text captions appeared in tandem. Astra said, adding a cache between the server and database could improve speed, end quote. Being out in the real world and being able to get answers to basic questions like, what am I looking at? I think you can see would be an obviously interesting use case that was never possible before. Ilius Soskever officially announced that he will be leaving open AI to work on what he called a personally meaningful project. Director of research, Jacob Pachaki will become OpenAI's chief scientist. Quoting CNBC, I am excited for what comes next, a project that is very personally meaningful to me, about which I will share details in due time. Suskever wrote in an ex post on Tuesday.

The departure comes months after OpenAI went through a leadership crisis involving co-founder and CEO Sam Altman. On Tuesday, Altman shared his thoughts on Susceiver's departure, quote, This is very sad to me. Ilya is easily one of the greatest minds of our generation, a guiding light of our field and a dear friend, Altman wrote on X. His brilliance and vision are well known. His warmth and compassion are less well known, but no less important, end quote. Altman said research director Jacob Pachocki, who has been at OpenAI since 2017, will replace Susceiver as chief scientist, end quote.

There's more. Jan Leek, who was co-leading OpenAI's super alignment team with Susceiver to steer and control more powerful AI has also resigned from the company. Look, I don't know if we'll ever know what exactly went down with all of this, as Narit Weiss Blatt pointed out. Six months this came after the failed coup, or specifically 179. days, to be exact, since the coup happened. This was clearly the end result of all of that, but why string it out this long? Why not part ways with Ilya earlier? And in the end, well, this summary from Crypto McKenna summarizes it perfectly, there's still so much we don't understand about all this. Quote, why would the chief scientist at Open AI leave when it's clearly

the central hub for the frontier? Why would Altman be removed as CEO and be reinstated later on? All of this is incredibly strange and suspect, end quote. This was most likely inevitable. Eight TikTok creators are suing the U.S. government over the law, forcing a sale or ban of TikTok in the U.S. arguing the move violates their First Amendment rights. Quoting the Washington Post, the law, quote, threatens to deprive them and the rest of the country of this distinctive means of expression and communication the creators suit states. It bans an entire medium of communication and all the speech communicated through that medium, even though at the very least the vast majority of that speech is protected, end quote. In a statement, the Justice Department said it looked forward to defending the

law in court, this legislation addresses critical national security concerns in a manner that is consistent with the First Amendment and other constitutional limitations. The statement said, the creators hail from eight states and represent a diversity of professions, backgrounds, and political leanings. They are Brian Firebaugh, 43, a cattle rancher in Hubbard, Texas, Chloe Joyce Sexton, 29, who owns a cookie baking business in Memphis, Kira Spann, a recent college graduate in Charlotte, who seeks to educate women about sexual violence and politics. Tofer Townsend, 33, an Air Force veteran and rapper who post videos quizzing people about the Bible. Talia Cadet, 34, a lifestyle creator in D.C., Timothy Marin, 25, a college football coach in North Dakota

who talks about sports. Paul Tran, 43, a beauty creator in Atlanta who owns a skincare brand, and Steve King, 50, a creator in Arizona who posts about comedy, fashion, and relationships. All of them told The Washington Post that TikTok is an essential outlet. let for speech and self-expression and that they rely on it to educate, advocate, and connect with millions of people, as well as to make a living. As Americans, we should be free to choose whatever app we want to use, especially when they haven't proved or provided evidence to show the danger they've set out that is unique to TikTok. Townsend said in a video posted to his TikTok account, end quote. Finally today, interesting Ray's humanity protocol is a blockchain-based

identity system that uses scans of people's palms to identify them, and it has raised a 30, million dollar seed led by Kingsway Capital at a $1 billion valuation. They want to go after projects like Sam Altman's World Coin. But again, let me stress the numbers here. $30 million on a $1 billion valuation. They achieved unicorn status only by giving away 3% of their equity. I'm not saying crypto is back or anything. I'm just saying crypto is the only place that I'm aware of where you can still raise a seed and immediately be a unicorn. Even in AI, that's not happening. quoting Bloomberg. Kingsway Capital led the seed funding round, which also included blockchain.com and schema capital as investors. Founder Terrence Quack said that the company also

raised about one and a half million dollars from influential crypto figures in a, quote, key opinion leader round. Founded late last year, humanity protocol plans to use the fresh capital to expand its more than 20-person team and build more partnerships as interest grows in platforms that confirm whether a person is a real person online. Quak said the startup plans to launch its test network during the second quarter and currently has a wait list of about 500,000. As it rolls out the technology, the plan is to first release an app that can use a phone camera to scan people's palm prints to determine their identity, Kwok said. The next step will be to roll out another layer of security that relies on

the network of veins in a person's palm and requires a small infrared camera that can connect to a smartphone. This system could help financial platforms conduct know-your-customer-customes or even be used to allow entry to physical locations, among other uses, according to Kwok. In the future, you can check into a hotel through a palm scan. he said, you can access your office building through a palm scan. The company also has plans to release a yet-to-be-named crypto token that can be used to pay verification fees. Both of its recent funding rounds were conducted through simple agreements for future tokens known in the industry as SAFs. Humanity Protocol isn't the only crypto project looking to combine biometrics and blockchain.

WorldCoin, which was co-founded by OpenAI chief executive officer Sam Altman, uses a device called an orb to scan people's eyes and give them a digital identity that demonstrates proof of personhood. However, the project has faced numerous regulatory setbacks around the world amid data privacy concerns as well as security issues. Quak said the use of biometric data has matured in recent years, pointing to the use of Apple's implementation of face ID in its iPhones. He said that humanity protocol system is not as invasive. The whole concept of identity is starting to become more important, Quak said. We look at artificial intelligence. We look at all the deep fake videos that are coming on.

It's not your face and it's also not your eyeballs, he said. It's much less dystopian, end quote. nothing more for you today talk to you tomorrow