One Manβs Quest to Reforest the Rio Grande Valley
Besides 2.0 Experimental Advanced mysteriously appearing (and disappearing) earlier today, a bug with the Gemini app has resulted in users losing the ability to create Audio Overview podcasts.
moreβ¦A new report reveals that the prototype Android XR glasses that Google recently showed off are slated to be released as a Samsung product as soon as next year, a bit later than other reports have hinted.
moreβ¦
For a while, it looked like President Donald Trump was going to have Big TechΓ’ΒΒs back.
Now, the tech industry is collateral damage in his global trade war.
On Thursday, European Commission President Ursula von der Leyen floated the idea of placing Γ’ΒΒa levy on the advertising revenues of digital servicesΓ’ΒΒ if tariff negotiations with the US go south. This would be the opposite outcome that tech CEOs like Mark Zuckerberg were hoping for when they threw their support behind the new administration.Β
To someone like Zuckerberg, Trump was supposed to be the strong-armed leader to bring the overbearing EU to heel. Instead, the rhetoric between the US and EU is ratcheting up just weeks before the EU is already set to fine Meta (and Apple) for violating its Digital Markets Act.Β
While certainly more of a self-inflicted wound, Elon MuskΓ’ΒΒs popularity in the US has Γ’ΒΒinverted as his support for President Trump has increased,Γ’ΒΒ Nate Silver wrote this week. TeslaΓ’ΒΒs stock price, meanwhile. has lost over a third of its value this year, and, thanks to tariffs, the company has removed the option to buy new, US-made vehicles in China.Β
As I predicted last week, TikTok is particularly s …
Netflix is starting to test search thatβs powered by OpenAI, according to Bloomberg.
The new search engine will let users βlook for shows using far more specific terms, including the subscriberβs mood, for example, the company said,β per the report. This OpenAI-powered search will also allow users to make queries that βgo well beyond genres or actorsβ names.β
The feature, which is opt in, is already available for some users to try in Australia and New Zealand on iOS.
Netflix spokesperson MoMo Zhou confirmed to The Verge that Bloombergβs story is accurate. Zhou says that the test will expand to the US βin the coming weeks and monthsβ and that there arenβt currently plans for the feature outside of iOS.
βItβs early days for the feature and weβre really in a learn and listen phase for this beta,β Zhou says.
OpenAI didnβt immediately reply to a request for comment.
In an interview on the Decoder podcast last year, The Vergeβs editor-in-chief, Nilay Patel, asked Netflix co-CEO Greg Peters about how the company was thinking about AI. Here was part of his response:
We have a long history of using machine learning and artificial intelligence in our recommender systems. Weβve been doing that for 20-some years. Again, we think that our job is to be proactive about understanding where thereβs technical innovation. How do we use that both to serve creators, allow them to tell their stories in more compelling ways, and also then to serve our members better user experiences?
There are few areas where AI has seen more robust deployment than the field of software development. From "vibe" coding to GitHub Copilot to startups building quick-and-dirty applications with support from LLMs, AI is already deeply integrated.
However, those claiming we're mere months away from AI agents replacing most programmers should adjust their expectations because models aren't good enough at the debugging part, and debugging occupies most of a developer's time. That's the suggestion of Microsoft Research, which built a new tool called debug-gym to test and improve how AI models can debug software.
Debug-gym (available on GitHub and detailed in a blog post) is an environment that allows AI models to try and debug any existing code repository with access to debugging tools that aren't historically part of the process for these models. Microsoft found that without this approach, models are quite notably bad at debugging tasks. With the approach, they're better but still a far cry from what an experienced human developer can do.