11-07-2025

Perplexity's $400M Deal Powers AI Search in Snapchat

Plus: Apple’s Siri upgrade with Google Gemini, diffusion model startup funding, and Laude Institute’s AI grants

Good morning, {{first_name | AI enthusiast}}. Perplexity is set to integrate its AI search engine into Snapchat’s My AI chatbot, backed by a $400 million deal that will expose the platform’s 940 million users to enhanced AI-powered queries starting early 2026.

This collaboration not only signals a strong commercial bet on AI search in social media but also raises questions about how AI-driven search will change user interaction and revenue models within high-traffic apps. Will this become the new norm for embedding AI capabilities into everyday digital experiences?

In today’s AI recap:

  • Perplexity’s $400M deal powering AI search in Snapchat
  • Apple’s Siri upgrade with Google’s 1.2T-parameter Gemini model
  • Inception’s $50M funding for diffusion models in code generation
  • Laude Institute’s Slingshots grants advancing AI evaluation research

AI Developments Spotlight: Perplexity's $400M Deal Powers AI Search in Snapchat

Editor's Comments:
Perplexity’s $400 million deal with Snap signals a major step in embedding AI-powered search directly into social media apps. Professionals and innovators should watch how this integration transforms user engagement and opens up new revenue streams in AI-driven interactions.

— Editor, BDCbox

The Recap:
Perplexity will pay Snapchat $400 million in cash and equity to integrate its AI search engine into Snapchat’s chatbot, My AI, exposing more than 940 million users to enhanced AI search starting early 2026. Read more.

Unpacked:

  • This deal places Perplexity’s search engine at the heart of Snapchat’s user experience, powering answers to questions through the My AI chatbot for over 940 million users worldwide.
  • Perplexity’s $400 million payment—split between cash and equity—reflects the growing commercial value companies see in embedding AI, with Snap expecting to record revenue starting in 2026.
  • Snapchat plans to roll out the AI search integration early next year, potentially setting a new standard for how social media apps automate information discovery inside their platforms.

Bottom line:
The collaboration highlights how AI tools are becoming integral to high-traffic consumer apps, setting the stage for professionals to harness AI-enhanced search and automation. This partnership marks a significant moment where AI's role expands beyond standalone tools into everyday digital experiences.

Apple’s Siri to Boost Productivity with Google Gemini AI Upgrade

Editor's Comments:
Apple is teaming up with Google to power Siri using the mammoth 1.2-trillion-parameter Gemini AI model, aiming to bring smarter, more capable voice assistant features in early 2026. This collaboration highlights a unique moment where top tech rivals join forces to enhance the AI tools professionals will rely on daily.

— Editor, BDCbox

The Recap: Apple is finalizing a $1 billion annual deal to run Google’s Gemini model on its private cloud for the next-generation Siri assistant, supercharging its reasoning and multi-step task handling abilities. The upgrade targets a launch in early 2026 to close the gap with competitors.

Unpacked:

  • The partnership gives Siri access to Google’s massive 1.2-trillion-parameter Gemini model, a vast leap from Apple’s current 150 billion-parameter baseline, promising more natural conversations and smarter follow-through. Digital Trends
  • Apple hosts the Gemini model on its own Private Cloud infrastructure, ensuring user data stays under Apple’s control while harnessing advanced AI power from a trusted source.
  • With this boost, Siri could soon manage complex commands like planning travel, booking hotels, and listing local spots all in one seamless flow, helping users save time and stay organized.

Bottom line:
This deal marks a big step toward making Siri a genuinely helpful AI assistant for everyday productivity. Professionals can look forward to Siri becoming more responsive and capable, reshaping how we interact with our Apple devices soon.

AI Developments: Inception’s $50M to Advance Diffusion Models for Coding

Editor's Comments:
Inception’s $50 million funding round highlights a promising new direction for AI in software development, focusing on diffusion models to improve speed and efficiency. Professionals and developers should watch this shift closely, as it may unlock faster, less resource-intensive code automation tools.

The Recap: Inception, led by Stanford professor Stefano Ermon, secured $50 million to develop diffusion-based AI models that generate code and text through iterative refinement rather than sequential prediction. These models power their Mercury software and already integrate into popular developer tools. Read more.

Unpacked:

  • Diffusion models work differently than autoregressive ones like GPT-5 by iteratively refining outputs instead of predicting word-by-word, offering faster and more parallelizable processing.
  • Inception’s Mercury model is already embedded in tools like ProxyAI and Kilo Code, making advanced AI assistance accessible for developers to speed up coding and reduce manual effort.
  • The approach delivers over 1,000 tokens per second in response speed, dramatically cutting latency and compute costs compared to traditional text AI models.

Bottom line:
This funding boost and technology pivot underline diffusion models’ growing role in practical AI applications, especially in coding. Expect more tools leveraging this approach to emerge, helping developers automate and accelerate complex programming tasks.

Laude Institute’s Slingshots Grants Accelerate AI Evaluation Research

Editor's Comments:
The Laude Institute has launched its inaugural Slingshots grant program to advance critical AI evaluation research and benchmarks. This funding and engineering support empowers projects that can shape the future of AI tools — an essential read for professionals tracking the next wave of productivity-enhancing AI innovations.

The Recap: The Laude Institute awarded 15 AI research projects with Slingshots grants, providing funding, compute power, and product support focused on new ways to evaluate AI capabilities and measure progress reliably. Learn more about the initiative here.

Unpacked:

  • The program supports projects tackling the notoriously difficult task of AI evaluation, aiming to ensure benchmarks keep pace with rapidly advancing AI models.
  • Well-known projects like Terminal Bench, a command-line coding challenge, demonstrate how these grants back efforts that assess AI coding skills in practical, developer-focused ways.
  • The grants emphasize core third-party benchmarks to maintain independent, reliable evaluation across AI tools, helping prevent benchmarks from becoming company-specific and less useful over time.

Bottom line: Slingshots grants funnel vital resources into creating benchmarks that validate AI progress, directly supporting the development of tools professionals rely on to boost their workflows. This initiative promises to sharpen the accuracy and usefulness of AI assessments, aiding smarter automation and innovation.

The Shortlist

Laude announced its first batch of Slingshots grants, funding 15 AI projects focused on evaluation benchmarks and reliable progress metrics, helping maintain independent testing standards for AI tools.

Sam reported that OpenAI reached a $20 billion annual run rate and plans $1.4 trillion in data center commitments over the next 8 years, emphasizing its massive scaling efforts in compute infrastructure and upcoming business verticals.

Amazon launched Kindle Translate, an AI-powered translation service aiming to broaden e-book authors' audiences, currently supporting English-Spanish and German-English translations, with a free beta to help indie authors expand global reach.

Sora saw nearly half a million installs on its Android launch day, over four times the size of its invite-only iOS debut, marking strong initial demand for OpenAI’s AI-powered video creation app across multiple markets.

Meta rolled out Vibes, its AI-generated short-form video feed, to Europe, allowing users to create, remix, and share AI videos on a TikTok-style feed, despite mixed community reactions to the flood of AI-generated content.

Sam clarified that OpenAI does not seek government bailouts or guarantees for its data centers, emphasizing market-driven success and cautioning against picking winners, while supporting public AI infrastructure for national benefit.

Get AI, CDP (Customer Data Platform) & SPA (Sales Process Automation) Tips

Get updates delivered
directly to your inbox.
2026 BDCbox© - All rights reserved

15-Minutes to 15 additional Sales a Month—Guaranteed.

Our conversational AI doesn’t just manage leads—it creates sales from your entire customer base. Let us show you how to squeeze more sales from your existing customer with a 15 minute, demo .

From Data to Deals: The Power of DIAA

Learn how Beatrix.ai transforms dealership data into personalized conversations that grow sales and service revenue. Schedule a 15 minute demo now.

15-Minutes to 15 additional Sales a Month—Guaranteed.

Our conversational AI doesn’t just manage leads—it creates sales from your entire customer base. Let us show you how to squeeze more sales from your existing customer with a 15 minute, demo.
  • Cass
  • NCOA
  • EmailValidation
  • EmailAppend
  • Mobile Phone Indicator
  • Mobile Phone Validation
  • Mobile Phone Append
  • VIN still owned and maintained by the customer