Meta is Poaching OpenAI’s Talent

It’s not just another weekend again, but the last one in June. The end of Q2

I know you think about it too. The year is running by so fast that you’d almost lose track of time if you shut your eyes for a little moment. You know how we all had new year plans at the beginning of the year? Some of you might not have taken it seriously but the tech giants have not rested since the year started. The competition is alot steeper now, and what that means for end users is that you get a a variety of products to try out, new features every new week and so much tech to catch up with.

For instance, this week, Google released a new product called Doppl that allows you visualize what your outfit will look like even before wearing it. We are telling you here first because we know so well that once this news gets out to the TikTok community, it’s over!

As usual, we are going to help you catch up on all the news you might have been this week. Enjoy!

Meta is reportedly in talks to buy a voice-cloning startup called Play AI, according to a new report from TechCrunch. The deal isn’t final yet, but if it goes through, Meta would be getting its hands on some pretty advanced tech that can clone voices and make AI-generated speech sound incredibly real. Play AI is still a relatively young company but has already raised over $23 million from investors like 500 Startups and Kindred Ventures. Its voice tech has been gaining attention, especially for things like AI-powered customer support where realistic voice interaction matters.

What makes this interesting is that Meta isn’t just trying to acquire the tech, they also want to bring on some of Play AI’s team. The goal seems to be boosting Meta’s efforts in building more human-sounding AI assistants, probably across apps like WhatsApp, Facebook, and even its smart glasses. This move follows a bigger trend. Meta’s been on an AI shopping spree lately. Just a few weeks ago, they invested over $14 billion in Scale AI, and they have been poaching researchers from OpenAI with some massive salary offers. Clearly, they are trying to level up in the AI race. If this deal with Play AI goes through, it could mean we will soon see smarter and more natural-sounding voice features built into Meta’s products.

Google just dropped a new app called Doppl that lets you see how clothes look on you without trying them on in real life. You upload a full-body photo, either a real one or an AI-generated mode, and then pick an outfit image from your phone, a website, or social media. Doppl uses AI to slap that outfit onto your photo, and it can even make a short video showing how the clothes move on your body. It’s free, only in the U.S., and available on both iOS and Android throughout June 2025. Google warns that it’s still experimental, so colors or fit might not be perfect, but it’s a fun peek at how AI could change shopping.

Early testers say Doppl is fun but a bit glitchy. For example, pants might not always fit well, socks can disappear, and mirror selfies sometimes turn people into cartoonish versions of themselves. Still, it has safeguards, you can’t use it to try on skimpy outfits or photos of celebrities without permission. Overall, Doppl is Google’s latest move into experimental AI fashion tech, building on earlier virtual try-ons in Google Shopping. It gives you a more personal look by letting you experiment with different styles, and share your favorite try-ons, though it’s still early days

Meta has brought on Trapit Bansal, a top researcher from OpenAI, to work on its next-generation AI reasoning models. Bansal was a major force behind OpenAI’s first logic-driven model, known as “o1,” joining the company in 2022 and contributing to its reinforcement learning work.

He will now be part of Meta’s new “AI superintelligence” team, a growing group led by the head of Scale AI, Alexandr Wang, and including recent hires like former OpenAI and DeepMind researchers. Meta’s goal is to build a powerful reasoning engine that can rival top-tier models from OpenAI and others. This hire comes amid Meta’s aggressive talent push, which has included recruitment offers rumored to reach $100 million, the recent addition of three ex-OpenAI researchers, Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai, and a major $14.3 billion investment in Scale AI. Meta doesn’t yet have a publicly available reasoning model, but adding top experts like Bansal gives them a good shot at catching up. These kinds of models use extra computation to break down complex problems before delivering answers, a capability that OpenAI and Google DeepMind have already deployed successfully.

Apple has just rolled out the second developer beta of iOS 26, and it brings a much-needed fix to the much-talked-about “Liquid Glass” Control Center. The first beta featured a translucent, glass-like interface, but users complained that it made the Control Center hard to read because it let too much of the home screen show through.

With Beta 2, Apple has increased the blur and opacity behind the Control Center, making buttons and sliders clearer and more user-friendly. While some color bleeding remains, it’s a solid step forward that addresses the biggest pain point users flagged. Reddit users noticed the improvement instantly, noting it’s “much better” even on busy home screens. Beyond the Control Center tweak, the beta brings a wider suite of updates: deeper Liquid Glass integration across iPad, Mac, Watch, and Apple TV; a fresh ringtone; enhancements to CarPlay; and early previews of features like live translation in chats and calls. These refinements show Apple is listening as it prepares for the public beta rollout in early July and the full iOS 26 launch alongside the iPhone 17 this fall.

Transform Your Ideas into Stunning Videos with AI. Turn Images & Text into Engaging Videos Instantly.