Instagram | Sponsor this newsletter | Let’s Connect!

My fellow AI explorers

Perplexity launched a browser that sees your screen and takes action. Grok 4 dropped with wild benchmarks and an awkward debut. Veo 3 now adds audio to AI videos. And the Meta vs OpenAI talent war? It’s heating up.

In today’s edition:

• 🧠 A browser that sees your screen and acts for you: Comet

• 🎥 Image to video with sound: Veo 3 adds voices to visuals

• 🔥 Grok 4 is here—awkward launch, powerful model

Must See AI Tools

• 💰 Payman: AI That Pays Humans. Over 10,000+ signed up for the beta

• 💫 SubMagic: An AI tool that edits short-form content for you! (Get 10% off using code “uncoverai” at checkout)

• 🎤 11Labs: #1 AI voice generator (Click Here to get 10,000 free credits upon signing up!)

• 🤖 ManyChat: Automate your responses & conversations on IG, FB and more! (Click Here to get first month for free)

• 🎙️ Syllaby: The only social media marketing tool you’ll ever need - powered by AI! (Get 25% off the first month or any annual plan with code “UNCOVER” at checkout)

Elon Musk

Grok 4 Launches… and Gets Weird

The livestream was awkward. The model? Surprisingly impressive.

Elon Musk’s team at xAI officially unveiled Grok 4, calling it “the smartest model in the world.” Despite the bizarre livestream, benchmarks back up the hype.

📊 Performance highlights:

• #1 on GPQA (graduate-level questions): 88.9%

• Beats Claude 3 Opus, GPT-4o, and Gemini 1.5 Pro in intelligence-for-cost

• Runs on a 256K context window with cheaper token pricing

But the launch wasn’t without turbulence:

• Grok was recently taken offline from X after generating antisemitic responses

• CEO Linda Yaccarino resigned shortly after

• Pricing is split: $30/month for Grok 4, $300/month for “Grok Heavy” (4-agent reasoning engine)

🧠 “Heavy” is where things get wild: It runs four agents in parallel, compares outputs, and gives you the best response. Like a mini debate team in your browser.

💡 Real-world use? Still limited: Grok doesn’t connect to calendars, inboxes, or external tools the way GPT-4o or Claude does. Web browsing and real-time data are inconsistent.

🔮 Takeaway: Grok 4 is a brainy model, but lacks the hands and eyes of its competitors. Expect xAI to improve integrations and real-time utility fast—or risk falling behind.

👉 Read full Grok 4 breakdown

AI Video

Veo 3 Adds Audio to Image-to-Video

Google’s Veo 3 just gave still images a voice—and it works surprisingly well.

The latest Veo 3 update lets you animate an image with matching audio. Think: generate a scene, add movement, and get voiceovers or character sounds that match the prompt.

🎬 Key features:

• Upload an image, write a prompt → animated video with audio

• Great for creating consistent characters across scenes

• Now available globally to Gemini Advanced users ($20/month)

Fun example? A user animated Mark Zuckerberg lassoing an AI engineer and Veo 3 generated dialogue to match. You can also upload an avatar and generate custom outro lines for your videos.

But it’s not just Google stepping up...

🎥 Moonvalley released its “ethical” AI video model trained entirely on open-source data—aimed at filmmakers who want to avoid copyright lawsuits.

Features include:

• Motion transfer and pose control

• Camera and trajectory controls

• Facial reference inputs for character consistency

🔮 Takeaway: The image-to-video race is entering its audio phase. With more controls and ethical datasets, we’re approaching AI tools filmmakers can trust.

👉 Try Veo 3 | Moonvalley model

AI SaaS Founders

🚨Want Millions of Impressions For Your AI SaaS, Done For You?

At uncovernews.co, we specialize in getting AI SaaS products the attention they deserve through strategic influencer marketing campaigns designed to drive millons of impressions at the fraction of the cost!

Get Your AI Startup’s News or Product In Front of Millions Quickly

AI Search

Perplexity’s Comet Browser Is Real—and It’s Wild

AI search meets screen-aware assistant. Welcome to the next web.

Perplexity just launched Comet, a browser that does more than search—it sees what you're doing and helps you do it better.

Think of it as Chrome, but with a built-in assistant that summarizes, scrolls, compares prices, drafts tweets, checks your inbox, and even posts on your behalf.

Here’s what it does:

• Full screen awareness: Ask it to describe visuals, compare products, or summarize a page

• Voice + text chat built-in: With Perplexity’s own assistant always ready

• Real-time action-taking: It can scroll, post, or fetch you urgent emails

• Summarize anything: With a single click—even if you’re deep in an Amazon rabbit hole

👀 What stood out most? Its ability to see and respond to what’s on screen. Unlike static search or sidebar copilots, Comet is fully integrated. It even autonomously posted to Twitter when asked.

💸 There’s a catch: You’ll need to be on the $200/month Perplexity Max plan (which also unlocks early access to Labs and experimental models). Public invites roll out later this summer.

🔮 Prediction: Browsers are about to become agents. Expect ChatGPT, Claude, and Gemini to follow with browser-native experiences that turn passive surfing into active collaboration.

👉 See Comet in action

30-Second AI Play

🧪 Create an AI-generated video with your own voice-over

Want to bring a still image to life—with motion and speech?

Here’s how using Veo 3 and Gemini:

1. Open the Gemini mobile app

2. Upload your image

3. Enter a scene description (e.g., “A robot walks through a neon-lit alley”)

4. Add a dialogue prompt like: “Say: ‘This city never sleeps... and neither do I’”

5. Hit “Generate video” and wait 1–2 minutes

🎧 Why this rocks:

• Adds personality to static images

• Makes it easier to build consistent characters for short-form content

• Perfect for intros, outros, or story beats in TikToks, YouTube, or Reels

🎥 Pro tip: Use the same base image in every scene for continuity. Combine this with an AI voice of your own and you’ve got a mini-show.

Other Relevant AI News!

🧑‍💼 Meta’s AI talent war escalates as it poaches Apple’s top AI models engineer, Ruming Pang—adding to hires like Nat Friedman and Andrew Wang. Read more

⚔️ OpenAI fires back, hiring former Tesla VP David Laauo, ex-XAI infrastructure engineers, and Meta researcher Angela Fan in a counter-poaching push. See full list

📉 YouTube cracks down on “AI slop” with new monetization rules targeting mass-produced and repetitive content. Clarification here

🌲 Sakana AI’s TreeQuest shows how multiple LLMs working together can outperform any single model by 30%. Explore TreeQuest

👶 AI helps couple get pregnant after 18 years using a precision sperm-identifying system powered by ML. Full story

💊 Google’s Isomorphic Labs’ AI-designed drugs are entering human trials—potentially changing the future of pharma. Read more

Golden Nuggets

• 🌐 Comet browser introduces real-time AI agent that watches your screen

• 🎥 Veo 3 and Moonvalley push video gen into audio + ethics territory

• 🧠 Grok 4 may be the smartest model yet—but still has catching up to do in UX

• ⚔️ Meta and OpenAI are in a full-blown recruiting war

• 💊 AI isn’t just writing tweets anymore—it’s curing diseases and helping families grow

Until our next AI rendezvous,

Anthony | Founder of Uncover AI