Updates and Insights for Today

1. Meta Snaps Up Voice AI Star Play AI!

2. Google DeepMind Unleashes ‘GenAI Processors’!

3. China’s Moonshot AI Drops Mega Model Kimi K2—Open-Sourced to Reclaim Glory

4. The latest in AI tech

5. AI Tutorials: Unlimited private web-scraping for AI Agents.

6. AI tools to checkout

AI News

Meta Snaps Up Voice AI Star Play AI!

Quick Summary:
Meta has officially acquired the voice-tech startup Play AI, bringing its full team and voice-generation tools in-house. This move bolsters Meta’s AI roadmap—integrating voice tech into Meta AI, AI Characters, wearables, and more.

Key Insights:

• Play AI specializes in natural, human-like voice generation and voice cloning tools.

• Meta confirmed the full Play AI team joins next week, led under Johan Schalkwyk in Meta’s AI division.

• The deal supports Meta’s broader AI effort—over $65B in 2025 capex, plus hires from OpenAI and Scale AI integration.

• Enhances Meta’s capability across voice interfaces in apps, devices, and content platforms.

Why It’s Relevant:
Meta is raising the stakes in the AI race. By folding Play AI’s voice tech into its ecosystem, Meta edges closer to conversational, voice-enabled services that rival rivals—and deepen immersion in the metaverse. Voice interfaces are gaining ground; Meta wants a leading role.

📌 Read More: TechCrunch, PlayAI

Our Partner Today

Make your marketing less boring

The best marketing ideas come from marketers who live it.

That’s what this newsletter delivers.

The Marketing Millennials is a look inside what’s working right now for other marketers. No theory. No fluff. Just real insights and ideas you can actually use—from marketers who’ve been there, done that, and are sharing the playbook.

Every newsletter is written by Daniel Murray, a marketer obsessed with what goes into great marketing. Expect fresh takes, hot topics, and the kind of stuff you’ll want to steal for your next campaign.

Because marketing shouldn’t feel like guesswork. And you shouldn’t have to dig for the good stuff.

Sign Up Free

Google DeepMind Unleashes ‘GenAI Processors’!

Quick Summary:
Google DeepMind just dropped GenAI Processors, a sleek open-source Python library that streamlines building fast, parallel, and real-time AI pipelines. It’s a game-changer for creating multimodal apps using Gemini and other LLMs.

Key Insights:

• The library provides a modular Processor interface to handle streaming of text, audio, video chunks seamlessly.

• It leverages Python's asyncio for asynchronous, concurrent execution—cutting latency and boosting throughput.

• Includes built-in processors like GenaiModel and LiveProcessor for real-time & turn-based workflows.

• Released under Apache‑2.0 on GitHub; ready to pip install and start innovating.

Why It’s Relevant:
GenAI Processors tackles a long-standing developer headache: juggling asynchronous pipelines with LLMs. Now, you can chain multimedia inputs (like live audio/video) through GPT-style models with minimal code and low latency. This could spark a wave of advanced AI apps—instant translators, live podcast analyzers, and next-gen assistants.

📌 Read More: Google Blog, GitHub

China’s Moonshot AI Drops Mega Model Kimi K2—Open-Sourced to Reclaim Glory

Quick Summary:
Moonshot AI, a Chinese startup founded in 2023, has released Kimi K2, a 1 trillion-parameter open-source model using a mixture-of-experts (MoE) architecture (32B active). The move comes as its Kimi app slid from 3rd to 7th in active users, and marks a bold bid to regain domestic AI prominence.

Key Insights:

• Kimi K2 uses MoE: 1T params total, 32B active, excels at coding, agentic tasks, and tool integration.

• Public benchmarks show it outperforms open models like DeepSeek V3 and rivals U.S. models on coding and software-engineering tests.

• Released under open-source license at launch, aligning with China's trend toward transparency to grow developer ecosystems.

  

  Orig: X

• On X (formerly Twitter), Moonshot shared: “Hello, Kimi K2! Open‑Source Agentic Model! … 1T total / 32B active MoE model … Strong in coding and agentic tasks”.

Why It’s Relevant:
Moonshot is leveraging openness as a strategic weapon. By open-sourcing K2 at release, they appeal to developers, undercut U.S. incumbents on price and flexibility, and fuel innovation via community. It's a defining shift as China’s AI players aim to dominate despite past chip restrictions and growing competition.

📌 Read More: X, VentureBeat

 AI Tutorials

Unlimited private web-scraping for AI Agents

Quick Summary:

Timothy Carbat, founder of AnythingLLM, demonstrates how the latest desktop release (v1.8.3) enables authenticated web scraping—allowing users to access and process private, password-protected content directly on their machines. This major upgrade empowers LLMs to finally “see” what lies behind login screens.

Key Insights:

• AnythingLLM is a local, privacy-focused AI app that runs on Mac, Windows, and Linux.

• It now supports authenticated web scraping—accessing logged-in content like LinkedIn or intranet data.

• All scraping happens locally and securely—nothing is sent to external servers.

• The tool integrates with LLMs to provide real-time, personalized responses based on private web content.

What Can I Learn?

• How authenticated scraping works using a built-in private browser

• Why most AI tools fail to access login-protected content

• How to integrate scraped data into agent prompts

• How to use the tool across systems for a truly local experience

Which Benefits Do I Get?

• Gain real-time AI access to personal data behind logins

• Improve context quality in LLM outputs

• Use on-device AI securely with no cloud dependence

• Seamlessly automate or personalize workflows with integrated agents

Why It’s Relevant:

Most LLMs today are locked out of the very data you care about—your inbox, dashboards, internal reports. AnythingLLM changes that. By enabling authenticated web scraping locally, it bridges the gap between powerful models and real, personal context. This feature redefines what AI agents can do for individuals and teams—privately, securely, and in real time.

Here is the full Video Tutorial 👉 Click Here

The latest in AI tech

1. Hugging Face’s Reachy Mini Charms as the ‘Seinfeld’ of Robots
Hugging Face unveiled the programmable Reachy Mini, an open-source, Raspberry Pi–powered robot with antennae and googly eyes. In its first 24 hours, it pulled in $500K in sales. The charm lies in its minimalism—not utility. It’s the whimsical equivalent of a Seinfeld episode: compelling without doing much.
📌 Read More: TechCrunch

2. Academics Sneak AI Prompts Into Papers to Skirt Criticism
Researchers in 14 institutions across eight countries are embedding hidden AI prompts in academic preprints, instructing LLM reviewers to give glowing reviews and overlook negatives. This tactic, revealed by Reuters and The Guardian, reflects tensions with “lazy” AI reviewers—raising ethical concerns about peer review integrity.
📌 Read More: The Guradian

3. McDonald’s AI Hiring Flaw Exposes 64M Applicants’ Data
McHire, McDonald’s AI hiring platform by Paradox.ai, had admin access protected by the password “123456.” Researchers accessed data of 64 million applicants—names, emails, phone numbers—within 30 minutes. The issue has since been patched, but it highlights critical risks in AI-powered hiring systems.
📌 Read More: CSO

4. Goldman Sachs Tests ‘Devin’—An Autonomous AI Coder
Goldman Sachs is piloting Devin, an autonomous coding agent by AI startup Cognition. Devin will assist—or replace—hundreds of entry-level software engineers among the bank’s 12,000 devs. This milestone signals growing AI integration into financial coding operations, sparking debates over human roles.
📌 Read More: CNBC

Our Second Partner Today

Start learning AI in 2025

Keeping up with AI is hard – we get it!

That’s why over 1M professionals read Superhuman AI to stay ahead.

• Get daily AI news, tools, and tutorials

• Learn new AI skills you can use at work in 3 mins a day

• Become 10X more productive

Sign up and start mastering AI

 AI Tools to check out

1. bldbl: Build your SaaS faster withMCP-powered tasks.

2. Clemence: Fully sourced data, freshly queried on the web.

3. Outfit: Get personal recommendations for places to go and stay.

4. Perso: Create, Translate & Launch Hundreds of Dubbed Videos Instantly!

Thanks for sticking with us to the end!

We'd love to hear your thoughts on today's email!

Your feedback helps us improve our content

⭐⭐⭐Superb
⭐⭐Not bad
⭐ Could've been better

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Tom