Google’s big week was a flex for the power of big tech

First: Gemini 2. But frankly speaking, I have become a bit used to language model performance updates, even to the point of indifference. Or at least bordering on indifference. i want to see them Do something.

So for me, the cooler update comes in second place: Project Astra, which feels like an artificial intelligence from a futuristic movie. Google first showed off a demo of Astra at its developer conference in May, and it became a hot topic at the show. However, since demos provide companies with the opportunity to showcase their most polished products, it can be difficult to tell what is real and what is just shown for the audience. Nonetheless, when my colleague Will Douglas Paradise recently Must try it yourselflive and improvised, it largely lives up to the hype. Although he found problems with it, he noted that they were easily corrected. He called the experience “amazing” and said it could be a killer application for generating artificial intelligence.

Among other things, Weill noted that this week the CEO of Google DeepMind, the company’s artificial intelligence arm Demis Hassabis receive his in sweden Nobel Prize. What did you do this week?“

Even more impressive is the fact that Willow, Gemini, Astra and Veo are making progress that many would have considered impossible just a few years ago, or at least not within this time frame.

A common criticism of the tech industry is that it has a tendency to overpromise and underdeliver. The phone in your pocket proves it. So did the number of rides I took in Waymo’s self-driving cars this week. (Both are arriving faster than Uber’s anticipated wait times. Honestly, the ability to summon an Uber isn’t that far away yet!) While quantum technology is still a long way off, Willow’s announcement seems like a Remarkable progress; if not exactly a tipping point, then at least a real signpost on a long road. (Anyway, I’m still not entirely convinced about chatbots. They do offer novel ways to interact with computers and revolutionize information retrieval. But are they good for humans – especially given energy debtthis Use of copyrighted material In their training data, they may struggle to overcome tendency to hallucinateWait – that’s controversial, of course. yes Under debate. But this week’s announcements from Google and OpenAI really shocked me—full stop.

For all the necessary and long-overdue discussions about controlling the power of big tech, only having Google (or Apple, or Microsoft, or Amazon, or Meta, or Baidu, or any other giant) can do that.

All that said, I don’t want us to buy more gadgets or spend more time looking at screens. I don’t want us to become more physically isolated and only interact with others through electronic devices. I don’t want us to fill the air with carbon or fill the soil with electronic waste. I don’t think these things should be the price we pay for progress. There is no doubt that humanity would be better served if the tech industry focused more on eradicating poverty, hunger, disease, and war.

Yet every once in a while, amidst the ever-present wave of hype and nonsense in Silicon Valley, epitomized by the AI ​​gold rush of the past few years, there are moments that leave me in awe and amazement: what people are capable of achieving, I’m hopeful about our ability to actually solve bigger problems – if only because we can solve so many other dumber but extremely complex problems. This week has been one of those moments for me.

Now read the rest of the report

News

• Adoption of robotaxis Reaching a tipping point.