It’s been a big week for Google and OpenAI (the maker of ChatGPT), but also for the world as a whole. We saw the evolution of AI and now have a clearer understanding of how it could be used in the real world.
The answer has left me quite excited but equally nervous about its impact: AI works so well in some use cases shown during the various presentations that it is clear and obvious that it will replace some jobs. How many depend on various things, but it will have a marked impact, so make no mistake about that.
On Tuesday, Google dedicated its two-hour Google I/O keynote to its generative AI platform, Gemini AI. This announcement came 24 hours after ChatGPT announced its new GPT-4o platform. Between the two, we’ve seen what the future of AI is, and this is before expected announcements and implementations from Microsoft on Monday and Apple in a few weeks.
Here are the AI use cases shown off this week and where I think they’ll have the biggest impact. You can recap Google I/O as a whole by following my Twitter thread, where I was live tweeting all of Google’s announcements in real time!
Impact by House of Tech is a reader-supported publication. Subscribe and support us in our mission to showcase how technology is having an impact on the lives of everyday people.
Project Astra is a game changer
I am most excited about Project Astra's platform and implementation. I’ve written about the Gemini AI features that will transform your life over at Android Police (as long as Google delivers), but Project Astra has the potential to completely revolutionize life for many people who may be otherwise less able.
Building Gemini into a pair of glasses is a smart way to implement the new multi-modal features; I.e., it’s no longer about just using a voice or typing; now, AI can see what you see and answer questions in reaction to what it sees. It’s smart, and it’ll only get smarter from there. As covered in various hands-on posts, the demo is limited. Project Astra can only remember the contents of a 4-minute session, and there are privacy concerns as the data is uploaded to the cloud, but this is a fascinating start.
Google was ten years early with its Google Glass product., On display at Google I/O was a new pair of glasses that Google used to showcase Project Astra, but I’d like to see them announced as a consumer-grade hardware product. Either that or Google needs to ensure partners such as Ray-Ban and Meta can use Project Astra as a software platform for the next generation of the Ray-Ban Meta.
Check out this post from Taylor for more on Project Astra.
Gemini is your new favorite study buddy
Many of the demos during the keynote featured Notebook LM, Google’s AI tool for asking questions about a set of documents. New for I/O is the ability to generate audio overviews, which create a podcast-like discussion between two speakers. Google has recognized that people learn differently, and the audio overviews feature is ideal for those who learn by listening rather than reading.
It can also give tailored examples, with Google’s Josh Woodward verbally asking Notebook LM, “Hey, can you give my son Jimmy a basketball example?” In just a few seconds, the two AI-generated presenters related the concept of force to basketball principles. Imagine your favorite lecturer always relating something to your interests. Now imagine being able to do that with absolutely any question.
Gemini will transform workspaces, for good or bad
Adding Gemini into Google Workspace will also have a major impact on millions of customers worldwide. We’ve already seen Microsoft integrate CoPilot into Windows in many ways, but if you’re a Google Workspace user, Gemini will completely revamp many of your simple workflows. Available via an extra charge - currently $20/m per user on an annual commitment or $24/m per user monthly - it adds the ability to do many things, including summarizing worksheets, creating complex workflows, and automatically moving items like receipts to your Google Drive.
The biggest impact will likely come from Google’s new AI teammate feature. Every business has wasted productivity from users asking questions like “How do I do ABC?”, “Was DEF approved?”, “Can we do XYZ?”. In-person, this causes minor disturbances as you can solve it in seconds, but in a remote team, this can end up meaning you disturb many people.
Gemini’s new AI teammate feature allows you to set up a new teammate that is all Gemini. Including Gemini in all the conversations and letting everyone in your team interact with it means it has access to your entire company’s knowledge base. This leads to being able to ask questions that would normally require a human response; the demo shown at I/O involved asking whether something was approved. Gemini was part of the conversation with two other team members who said it was approved, but the message hadn’t filtered back yet. The AI teammate could recall that information from the separate chat and highlight in the prompt’s answer that it was approved and by whom.
This is all good, but there is a fundamental other problem here: it’s clear that AI can replace many lower-level jobs and, presumably, as it gets smarter, higher-level jobs as well. The improvement in both Gemini and ChatGPT in the past six months is incredible and also showcases a frightening future for many people. Most employees have little job security — especially in the US, where everything is at will — and Gemini AI will only serve to reduce this further.
Read this from The Verge for more on Gemini for Google Workspace!
The other notable Gemini features
Spam detection will also make a big difference in daily life. I love that Google used many real-world examples to showcase the power of certain AI features, and Spam Detection will help everyday users. I can imagine that many people I know will benefit from this feature; although they’ll most likely be confused at first by the prompt, I can see this saving millions of people a lot of money.
Gemini can also recognize video and answer questions in real time based on what it sees through the camera. Hold your phone up, and it can understand what it sees and answer any questions you have. The demos looked fantastic and could be exponentially useful for many people. Google even referred to its Circle-to-Search feature — which launched with the Galaxy S24 series but has rolled out to other devices — as the Circle-to-Search Study Buddy, as it can now recognize video.
This ability for Gemini AI to see is going to transform the lives of Android users who are visually impaired. The updated TalkBack feature — an accessibility feature that allows your phone to give you visual prompts to guide you around actions or highlight what you’re doing on your phone — can now provide more detail about images on a screen, say of something you’re attempting to buy. No more relying on someone’s implementation of ALT texts (the textual description for an image manually created by the author), instead let AI do all the work and do it better. It looks fantastic and will be a major boon for any visually impaired users.
Check out the video below from The Verge that summarizes all of Google I/O in 17 minutes! It’s worth watching if you want to know more about AI!
Gemini is Google’s answer to ChatGPT-4o
Gemini may be the major subject of this piece, but it came in response to another breakthrough AI announcement 24 hours earlier. OpenAI revealed the new ChatGPT-4o, its most advanced AI model yet that is available to everyone.
Like Gemini, it is multi-modal and can recognize text, photos, and videos. The demo was super slick and showed two AI bots talking to each other in real time, with GPT-4o able to interact with the other bot and answer questions based on the content in the other bot’s viewfinder. There were also demos of GPT-4o being used to help folks with homework and tutor children on Maths.
AI will help to empower many people, especially those who want to help their children but may not have the prerequisite knowledge to do so, and it’s great to see that both Google and OpenAI have heavily focused on these use cases in their demos.
Check out this guide from Ryan at Tom’s Guide on ChatGPT-4o (for Mac).
Elsewhere in tech this week
I planned to do a separate summary about tech this week, but it’s all AI. Yes, we saw the iPad Pro go on sale finally, and while I’m working on the review, I’ll say this: the iPad Pro 2024 is an absolute game changer for the tablet space and portable productivity. It comes at an incredible price, though — starting at $999 but rising to over $2,100, including the Magic Keyboard and Apple Pencil Pro — but it may be worth it to some people. Stay tuned to my Twitter for more!
I wanted to write about many other things this week, but AI took over everything. I’m currently sitting in a hotel in Delaware, writing this on my iPad Pro before I fly to a couple of weddings. I’m on day 2 of a crazy few days of travel and tracking sleep and stress using three of my favorite wearables: Apple Watch Ultra 2, Galaxy Watch 6 Classic, and the Oura Ring. Stay tuned for lots more on all of these in the coming weeks!
Impact by House of Tech is a reader-supported publication. Subscribe and support us in our mission to showcase how technology is having an impact on the lives of everyday people!
Until next time, Happy Friday Saturday (I wrote this yesterday but only just got around to publishing!), and stay healthy, my friends!
