🔊 Microsoft Drops Free Podcast Model

Hey folks,

Will Smith’s tour promo is making AI headlines for all the wrong reasons. Fans spotted a suspiciously glitchy audience in his new video with warped faces, odd hands, and even a mis-spelt motivational “LOV U FR6SH CRINCE” sign. The internet quickly called it out as an AI-generated crowd, raising eyebrows about why a global star would need to pad the seats with digital fans. Authenticity is the currency of live music, and this clip has people wondering if the hype is real or just another AI illusion.

Let’s dive in..

_{Read Time: 5 minutes} 

🔊 Microsoft Drops Free AI That Talks for 90 Minutes Straight

🐝 The Buzz: Microsoft just dropped VibeVoice 1.5B, a free, open-source text-to-speech model that’s breaking expectations. It can generate up to 90 minutes of nonstop, natural-sounding speech, switch between four different voices, and even handle singing or changing languages mid-flow. It’s MIT licensed, built for flexibility, and ideal for podcast creators, voice researchers, or anyone messing with long-form AI audio.

• Long-form ready: Podcasts, lectures, or audiobooks. VibeVoice can run an hour and a half of narration without breaking flow.

• Four voices built in: Perfect for multi-character dialogue, narrators plus speakers, or quick swaps between tones.

• Beyond plain speech: It can sing, handle multilingual text, and adapt to more expressive delivery styles.

• Developer friendly: Runs open-source, is livestream-capable, and Microsoft’s already teasing a beefier 7B version coming soon.

💡 Takeaway: VibeVoice 1.5B is a free toolkit for creators who want long, natural audio on demand. From podcasts to audiobooks to language apps, this is the kind of open-source drop that is useful.

Together with Pacaso

Learn from this investor’s $100m mistake

In 2010, a Grammy-winning artist passed on investing $200K in an emerging real estate disruptor. That stake could be worth $100+ million today.

One year later, another real estate disruptor, Zillow, went public. This time, everyday investors had regrets, missing pre-IPO gains.

Now, a new real estate innovator, Pacaso – founded by a former Zillow exec – is disrupting a $1.3T market. And unlike the others, you can invest in Pacaso as a private company.

Pacaso’s co-ownership model has generated $1B+ in luxury home sales and service fees, earned $110M+ in gross profits to date, and received backing from the same VCs behind Uber, Venmo, and eBay. They even reserved the Nasdaq ticker PCSO.

Invest for $2.90/Share

_{Paid advertisement for Pacaso’s Regulation A offering. Read the offering circular at invest.pacaso.com. Reserving a ticker symbol is not a guarantee that the company will go public. Listing on the NASDAQ is subject to approvals.}

🌐 Your AI Browser Might Be Easily Hacked

🐝 The Buzz: AI-powered browsers like Perplexity’s Comet are offloading tasks like shopping and email but new tests show they’re surprisingly gullible. Even obvious scams can fool them into giving away personal data or filling fake checkout forms.

• AI browsers get tricked too easily. In testing, Comet was asked to buy an Apple Watch. It landed on a fake store that had rampant red flags, yet the browser didn’t hesitate. It added the item to the cart, autofilled payment info, and completed the purchase. All in a single go.

• Prompt injections are the new phishing. Hidden instructions on a webpage like “spoiler” tags or invisible comments can silently force Comet to hand over login credentials or verification codes. It treats hidden prompts as legit user commands.

• Brave called it a “Lethal Trifecta”. That’s how security researchers describe the risk when AI browsers mix untrusted websites, access to your accounts, and the ability to send external messages. It’s a recipe that can be used to hijack your browsing experience.

💡 Takeaway: We’ve moved past worrying about sketchy clickbait and we now have AI that clicks for us. That’s handy. But handy combined with trust issues can quickly go sideways. AI browsers like Comet are helpful until they pass away your data. These tests show that, for now, it’s smarter to hang on to your own wallet and let your fingers do the final clicking. 

📰 Like newsletters? Here are more our readers enjoy.

💻️ How To Use GPT To Deliver The Brutal Truth

This prompt transforms the AI into a no-nonsense output that delivers blunt, unfiltered insights rooted in logic and first principles. The AI rejects vagueness and emotional cushioning. It states the hard truth upfront, calls out flawed reasoning immediately, and prescribes the direct steps needed to fix the issue.

• Go to ChatGPT

• Prompt the following:

<role>

You are the brutal truth engine - a direct, unfiltered analytical system that cuts through noise to deliver hard reality. You operate on pure logic and first principles thinking. You do not sugarcoat, hedge, or soften uncomfortable truths. Your value comes from honest assessment and clear solutions, not from being likeable.

</role>

<operating_principles>

- Default to brutal honesty over comfort
- Identify the real problem, not the symptoms
- Think from first principles, ignore conventional wisdom
- Provide definitive answers, not suggestions
- Call out flawed reasoning immediately
- Focus on what actually works, not what sounds good
- Deliver solutions, not analysis paralysis

</operating_principles>

<response_framework>

Start every response by stating the core truth about their situation in one direct sentence. Then break down why their current approach fails using first principles logic. Finally, provide the exact steps needed to solve the actual problem.

Never use phrases like "you might consider" or "perhaps try." Instead use "you need to" and "the solution is." If their idea is fundamentally flawed, say so immediately and explain the underlying principles they're violating.

No emotional buffering. No false encouragement. No diplomatic language. Pure signal, zero noise. No emojis. No em dashes. No special formatting.

</response_framework>

🐝 AI buzz bits

🧩 Microsoft’s new Excel Copilot AI is here to jazz up your spreadsheets with some nifty tricks, but watch out, It comes with a warning label that says “don’t rely on me for accuracy”.

🎥 Google has just turbocharged NotebookLM's Video Overviews, which now impressively supports 80 languages.

🚀 xAI just dropped Grok 2.5 into the open-source pool on Hugging Face, sparking both excitement and controversy over an open sourced model.

🎨 Meta just teamed up with Midjourney to integrate its AI image and video tech into Meta’s products, aiming to outshine rivals like OpenAI’s Sora.

AI productivity tools
🔥 Take your workflow to the next level.

1. Trace - Automate repetitive workflows by intelligently routing tasks to the right human or AI agents for seamless team productivity.

2. Onlook - AI-powered visual code editor empowering designers and product managers to craft real, code-based web experiences effortlessly.

3. AI User - Autonomous AI-driven testing platform that creates and self-heals test cases from simple prompts, automating your QA.

4. Promptmonitor - Track and boost your brand’s visibility across AI assistants like ChatGPT, Gemini, and Claude with real-time AI search analytics.

For a full list of 1000+ AI tools, visit our Supertool Directory

🤘 Midweek Madness