Apple Paying $1 Billion for Google's AI Brain for Siri 💰

Welcome to another edition of Horizon AI,

With the urgency to catch up in the generative AI race, Apple is reportedly close to finalizing a major, multi-year deal to integrate a custom-developed AI model from its rival, Google, into its Siri voice assistant.

Let’s jump into it!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

• Siri's Brain Transplant: Apple Taps Google AI🍎

• Pantone's AI Palette: Microsoft Powers the Future of Color

• AI Tutorial: How to Create a Personalized Podcast with AI Voices Using AI

• AI Tools to check out

• AI Findings/Resources

• The Latest in AI and Tech 💡

AI News

APPLE

Siri's Brain Transplant: Apple Taps Google AI

Apple is reportedly planning a major overhaul of its digital assistant, Siri, by incorporating an AI model from its competitor, Google. This move marks a significant partnership between the two technology giants, with the aim of enhancing Siri's capabilities with more sophisticated features.

Details:

• Apple will pay Alphabet's Google about $1 billion a year for access to its AI technology and utilize Google's generative AI model to power the next iteration of Siri.

• The integration is expected to significantly enhance Siri's capabilities, particularly for tasks that require complex, conversational responses. Google's model will serve as a stopgap solution, providing the intelligence needed for the Siri overhaul until Apple's own in-house systems are sufficiently developed.

• The system will utilize a 1.2 trillion-parameter artificial intelligence model, a measure of the software's complexity that is vastly superior to Apple's current systems. While not explicitly named, the AI model is reportedly a version of Google's Gemini family.

• The model will run on Apple's Private Cloud Compute servers, ensuring user data remains walled off from Google's infrastructure.

The move is critical for Apple as it aims to integrate advanced AI features into the next version of its operating system, iOS 18, to remain competitive with other tech companies.

Source

TOGETHER WITH NATIVEMIND

Your AI, Your Device, Your Rules.

Meet NativeMind: a fully private, open-source AI that runs 100% on your computer. No cloud. No login. Nothing leaves your hands. It thinks fast, works offline, and fits right into your browser. Use it to write, translate, search, and handle email, without giving up privacy or speed.

Key highlights:

• Agent Mode handles multi-step tasks on-device.

• Local search finds answers inside your browser.

• Gmail help writes and summarizes securely.

• Chat with PDFs, images, and tabs.

You stay in control, always. Add open tabs as context. Ask across sites. Translate whole pages without breaking the layout. Soon, you can chat with files entirely offline. This is AI you can actually trust, built for people who value privacy and performance.

Install in seconds and get private AI superpowers today, free.

Get NativeMind now. Keep your data yours 🔒.

MICROSOFT

Pantone's AI Palette: Microsoft Powers the Future of Color

Pantone, the global authority on color, is partnering with Microsoft to integrate artificial intelligence into the creative design process. This collaboration aims to guide designers into the future by providing them with powerful new tools to enhance their craft using intelligent technology.

Details:

• Pantone unveiled the beta version of the Pantone Palette Generator, which is built using Microsoft Azure OpenAI and integrated into Pantone Connect, the company's digital platform.

• This new tool features a chat-based interface that allows users to ask for color palettes and design concepts based on questions like "What colors evoke optimism in Gen Z?"

• The tool uses semantic search and insights from the Pantone Color Institute to deliver instant, inspiration-driven color palettes, saving designers hours of research and speeding up the discovery phase.

• The palette generator is powered by Retrieval-Augmented Generation (RAG) and agentic technology, grounding its color suggestions in Pantone's proprietary trend forecasting data.

• The tool, accessible via Pantone Connect, integrates Pantone's proprietary color library and forecasting insights with AI-powered search and generation, providing creatives with a faster, insight-driven workflow while preserving Pantone's design heritage.

Source

AI Tutorial

How to Create a Personalized Podcast with AI Voices Using ElevenLabs

1. Once you have your theme and short script, Sign Up on ElevenLabs.io

2. Choose from built-in voices from the ElevenLabs library or upload a consented sample to clone your own.

3. Convert your script into SSML (Speech Synthesis Markup Language) to control pauses, pitch, emphasis, breaths, and intonation.

   • Example SSML snippet:

<speak>
  <p><emphasis level="moderate">Welcome to Episode 1.</emphasis></p>
  <break time="400ms"/>
  <voice name="your-voice">
    Hi — I'm Alex, and today we explore how AI is changing creativity.
    <break time="300ms"/>
    <prosody rate="0.95">Let’s jump in.</prosody>
  </voice>
</speak>

• Test short passages and tweak prosody, pauses, and punctuation until it sounds natural.

4. Paste SSML or plain text into the Studio or API and generate segments separately (intro, ad, each chapter, outro) so you can edit/mix later. You can use “voice variations” / “emotion” controls to add natural inflection.

5. Download the audio files, then use an editor like Audacity or Descript to trim, remove artifacts, or add background music, fade-ins, and outros where needed.

6. Save as MP3 and upload to a podcast hosting service (such as Spotify for Podcasters, Buzzsprout, or Podbean). You should also write a short episode description while sharing your link.

   ✅ Pro Tips

   • Keep episodes between 10–20 minutes.

   • Add natural pauses and enthusiasm for realism.

   • Update your voice or tone to match the episode’s theme.

AI Tools to check out

🛠️️MeDo by Baidu: Build production‑ready full-stack apps with AI in minutes.

🔗Context Link: Get better context with AI solely from shared Personal URLs or Docs.

⚖️GitLaw: AI tool to review and draft legal documents

📧Extra Thursday: AI email manager to draft, organize, and respond to emails faster.

🛡️️Snyk: AI cybersecurity solution to find, prioritize, and fix vulnerabilities

AI Findings/Resources

📚 365 Data Science is offering free AI, data science, and machine learning courses from beginner to advanced, with free certificates, for a limited time.

📊 2025 AI Index Report by Stanford HAI: a comprehensive look at AI progress, benchmarks, applications, and societal impact.

🌐 The State of AI: Global Survey 2025 by McKinsey & Company

🚗 NVIDIA open-sourced 1,700 hours of multimodal driving sensor data for “physical AI” applications (driving, robotics, sensor fusion).

The latest in AI and Tech

AMD Beats Q4 Forecasts, Fueled by Chip Demand

AMD forecasts Q4 revenue above estimates, driven by strong demand for AI chips, fueled by investments in data center infrastructure and partnerships with OpenAI and the U.S. Department of Energy.​

IBM's AI-Driven Layoffs

IBM announced staff reductions, specifically targeting roles that the company believes can be replaced by advancements in artificial intelligence. This strategic move signals that even major technology companies are prioritizing efficiency by using AI to automate tasks, leading to changes across their workforce.

Bank of America's Billion-Dollar AI Bet

Bank of America is aggressively ramping up its major AI projects after increasing its annual spending on new technology by 44% over the last decade. This intensified focus on larger AI initiatives is aimed at driving efficiency and competitive advantage across the entire enterprise.

Anyone Can 'Jailbreak' AI

Intuitive questions are effective at causing AI chatbots to produce biased or "jailbroken" responses, as experts using complex technical methods. Penn State's research reveals new challenges for AI safety as everyday prompts, not just sophisticated attacks, can exploit and reveal inherent biases in large language models.

AI Accelerates the Antibody Design to Battle Novel Viruses

Vanderbilt researchers have demonstrated that using a "protein language" AI model can rapidly design functional monoclonal antibodies against emerging viral threats like RSV and avian flu. This breakthrough is a step toward using computers to efficiently design novel therapeutics from scratch for a broad range of diseases, including cancer.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

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

See you in our next edition!

Gina 👩🏻‍💻