8 Things Google Announced at I/O

At its annual developer conference, the focus was on A.I. and how it will change the way we interact with, well, everything.

When most people think about artifical intelligence (A.I.), I suspect they are thinking about computers that are capable of thinking for themselves. Even though A.I. has become the trendiest of buzzwords, we’re still a long way from machines capable of fully independant thought and decision-making.

Instead, somewhere along the way the companies that make computers—and computer software—decided that the term could mean whatever they wanted, and so now it mostly just means “computers doing stuff.”

That doesn’t mean some of that “stuff” isn’t very cool. At Google I/O this week, the company announced a series of things that are both very cool and that will probably change the way you interact with computers.

Photos

Google Photos has always had computational photography as its main benefit. Now, with a feature called Magic Editor, you can use generative A.I. to edit your photos. So, if you didn’t quite get the framing right, Google Photos will reframe the image for you, and create parts of the image with generative A.I. to fill in the rest.

This isn’t just about filling in a new sky, or eliminating distracting elements from the background. In the demo, Google Photos recentered an image on the subject and generated a background. It begs the question “what even is a photo?”

Maps

Maps gets a new way to preview a route when asking for directions. Known as “Immersive View for Routes, you can now get a flythrough preview of the route you’re taking that uses Google’s satellite images.

If, for example, you were going for a bike ride, you could decide which route you prefer based on what you’ll see along the way. I imagine it would also be helpful if you’re in an unfamiliar city. You could preview the turn by turn directions as you’ll experience them. It’s limited to New York, London, San Francisco, and Tokyo for now, but if you happen to live or travel in one of those cities, it’s a cool way to see where you’re going.

Bard

Bard is Google’s chatbot, and it’s now available to everyone—there’s no more waitlist. Not only that, but Google says Bard will now integrate with Workspace. That means that if you use Bard to generate a list of items, you can send that directly to Google Sheets or Docs and it will know what you’re asking.

Duet

In addition to Bard, Google has brought the same kind of generative A.I. capabilities directly into Google Workspace and called it Duet. This seems like a direct answer to Microsoft Copilot, which allows you to take a Word Document of a marketing proposal, and turn it into a Powerpoint presentation.

Google is implementing the same type of capabilies into its own office productivity suite. The features have been available for a few months, but now they have a name.

Search

Search might be the biggest change of all, for two reasons. First, the internet is mostly organized around Google Search. It’s where the majority of people start their interactions online, and any change to the way search works affects literally billions of people.

The other reason is that Google Search is also one of the most effective and widely-used marketing tools. Businesses work hard to have their website show up in the list of blue links for a given search inquiry. Now, however, Google is building a more A.I.-centered search experience.

When you search for a given phrase, you’ll see an A.I. generated box at the top that will summarize an answer to your question, based on results across the internet. You can then ask a follow up question to refine the information.

Sure, Google says it will still show traditional links below that section, but the reality is that people simply don't scroll down below what they can see in their browser view. Also, interacting with an A.I. chat version of Google is far more engaging than scrolling through dozens or hundreds of links to websites--many of which only exist as SEO spam. 

It's one of the main reasons Google has been far more cautious in rolling out Bard, its A.I. chat feature. Google can't afford to get this wrong, both in terms of its reputation (A.I. chat robots tend to have a pretty loose relationship with the truth), and in terms of its business model. 

Pixel Fold

In addition to new features in existing products, there were a few new hardware devices. The first was the Pixel Fold, which Google itself had already sort of announced in a tweet ahead of time.

I only had a few minutes with the Pixel Fold, but my first impression is that it’s very well done. Compared to the Galaxy Fold, Google’s version feels just as premium. It’s clear Google has been working on both the hardware design challenges that come from making something with a screen that folds up, as well as tuning Android to work well on a device that is sometimes a phone, and sometimes a tablet.

It’s definitely the priciest Pixel ever at $1799, but Google does say that if you pre-order one, you’ll get a free Pixel watch.

Pixel Tablet

The bar for making an Android tablet is so low that Google didn’t have to do much to make something that isn’t terrible. In the world of tablets, Android-powered devies are almost all pretty bad. Partly that’s because people who want a tablet computing experience just buy iPads. Partly it’s because Android has never been good on a tablet form factor.

With the Pixel Tablet, Google has solved the problem by making what seems like a highly capable device running a version of Android that knows what it’s doing on a tablet. In addition, the Pixel Tablet has a very good party trick in that it comes with a docking speaker that turns it into a sort of home entertainment device.

In the hands-on area, the speaker sounded very good for something that size. It’s exactly the kind of thing a lot of people will love. It’s also something I wish Apple would make for the iPad.

Pixel 7a

The Pixel 7a is the affordable Pixel smartphone, and therefore, it’s the one most people are most likely to buy. For $499 you get a device with a good screen and a best-in-class camera system. I’ll have more to say when I have a chance to review one for myself, but if you’re looking for an affordable Android device, it certainly seems like a worthwhile option.

Other news

Twitter is getting a new CEO

In a Tweet, Elon Musk says he has found his replacement as CEO of Twitter, or X as the company is now known. You might remember that Musk promised in December to step down once he found someone "foolish enough to take the job." That came after a poll overwhelmingly said that he should resign. 

The Wall Street Journal is reporting that that person is Linda Yaccarino, the head of advertising at NBCUniversal. That would make sense considering Twitter is an ad business—though it’s never been particularly great at making money.

I covered this more for my column this morning at Inc.com.

Bill Gates was asked about A.I.

In an interview with ABC News, Bill Gates was asked about A.I., and specifically, fake images. In response, Gates suggested that everyone is responsible for figuring out what is fake and what is real.

"The reader, the viewer, every person has the responsibility to question,” Gates told Rebecca Jarvis. “Society is going to start to say, 'OK, this source is more of a trusted source.' It's not like we can say, 'Hey, please stop making fake images.'"

Gates is right that it's impossible to stop people from making fake images. That's been a thing since long before computers would do it for you all on their own.

The problem is, people generally aren't great at distinguishing whether something on the internet is from "more of a trusted source." Instead, a lot of people just look for whatever reinforces what they already believe. That's why misinformation online is so powerful. If you're predisposed to believe a thing, and then you see a photo of that thing, for example, it becomes very hard to filter through what is true and what isn't.

If nothing else, the political climate of the past six years has shown that just because information comes from what were once considered widely trusted sources, it doesn't mean everyone will believe it. The opposite is even worse. Even when information comes from sources that are obviously not credible, millions of people will believe it because it's what they want to hear, read, or see.