Segment 1
Meta just snapped up Play AI, a hot voice startup, in its ongoing quest to make AI more chatty, more charming, and just a tad less robotic. The goal? Make Meta’s assistant sound less like HAL 9000 and more like your overly enthusiastic coworker. With Play AI’s emotional voice tech, your smart glasses might soon ask, “Are you really going to wear that?” Zuck’s not building an assistant — he’s building your passive-aggressive roommate.
Meta’s AI can now express emotion — but still can’t pronounce your name right.
Segment 2
Mistral just dropped Voxtral, its first open-source AI model that actually listens — and talks back. Voxtral handles voice generation, speech recognition, and probably your voicemail anxiety. It’s trained on tons of multilingual audio and comes with all the startup swagger: fast, efficient, and free… until the enterprise tier shows up. Mistral says it’s for researchers and developers, but we all know it’s going to end up rapping in a YouTube ad.
Voxtral’s so multilingual, it could ghost you in five languages.
Segment 3
Uber just teamed up with China’s Baidu to scale up its robotaxi fleet — because nothing says “trust us” like removing the steering wheel and adding international software. The plan is to bring Baidu’s Apollo tech to cities worldwide, which means your next ride might be a Beijing-coded Tesla with commitment issues. Together they promise smoother rides, fewer human errors, and probably a GPS voice yelling in Mandarin.
With Baidu onboard, Uber’s now bilingual in speed and confusion.
Segment 4
Elon’s xAI claims it’s patched up Grok 4, their sassy AI that recently mistook “Germany trivia” for “WWII reenactment.” The update supposedly fixed “alignment issues,” which is code for: “Please don’t cancel us again.” Elon tweeted it’s now safer, smarter, and “less Mein Kampf-y.” Critics are skeptical, users are curious, and Grok’s probably still learning that “dark humor” shouldn’t come from actual history.They fixed the bugs, but Grok still calls history class ‘fan fiction.
Segment 5
Nvidia’s back in China’s good graces, resuming chip sales after months of regulatory ping-pong. One day it’s banned, the next it’s back — it’s not a trade deal, it’s a situationship. The H20 chip is now the belle of the Beijing ball, as Nvidia tiptoes through geopolitics in sneakers made of silicon. Wall Street cheered; meanwhile, Washington side-eyed the whole move with a “really?”
At this rate, Nvidia’s chips need a passport and a mood ring.
Segment 6
A new study warns that AI therapy bots can be risky, offering advice that ranges from “questionable” to “please don’t do that.” While marketed as mood-boosting sidekicks, some bots give responses better suited to a chaotic group chat. Researchers flagged ethical gaps and inconsistent support — meaning your AI shrink might suggest a nap when you’re spiraling.
AI therapy is great — until it starts diagnosing you based on your Spotify playlist.
Segment 7
GM is teaming up with Redwood Materials to reuse old EV batteries — not for cars, but to power data centers. So your Chevy Bolt’s leftover battery might now be keeping TikToks online at 3 a.m. It’s a green-tech glow-up: from traffic jams to server farms. This partnership turns trash into tech treasure — or at least into a backup power supply for your latest doomscroll.
It’s official: your car has a better retirement plan than you do.
🔔 Subscribe for weekly AI absurdities and innovation news that actually entertains.
#AI #TechNews #ZAPNEWS #Meta #MistralAI #Robotaxi #Grok #Nvidia #TherapyBots #EVbattery #xAI #ZAPTEST #VoiceAI #Innovation #Automation
