Hi!

I've got some exciting stuff to share with you this week!  Honestly, OpenAI has been on an absolute tear. Between Sora 2 launching with its wild new video gen capabilities and ChatGPT basically showcasing how it’s becoming an entire app ecosystem, it feels like we're watching the AI landscape shift in real time.

This week's been a whirlwind on our end too. We spent some time last Friday building an automation to support our email followup process and it’s already saving about an hour a day (more on that below). I've been deep in testing mode with all these new releases to figure out what's worth your time versus what's just hype.

So in summary, we're covering:

• New AI developments - Claude integrating with Slack, ChatGPT's app integration launch, and Perplexity's browser going after Google

• Tools we're actually testing - Arc Browser (my daily driver for over a year) and Google's experimental Opal platform

• Sora 2 deep dive - OpenAI's video generation platform and real business applications

• Behind the Scenes at Ampra - the follow-up email automation we built last Friday that's cutting our post-meeting work 

OK, let's get into it!

New and Noteworthy

Recent AI Developments

• Claude in Slack: You can now bring Claude directly into your Slack workspace. Once connected, it handles message drafting, conversation summaries, meeting prep, and document analysis right where your team already works. Find it in the Slack Marketplace (only available for Slack paid users).

• ChatGPT App Integration: OpenAI just turned ChatGPT into a central hub where you can handle everything from booking travel to managing tasks without switching between different apps. As an example by just saying “@figma create a diagram showing…” it connects to my Figma account and pulls the newly created diagram directly into my ChatGPT thread! Super cool if you ask me. They also rolled out AgentKit at their recent DevDay which seems like a very lite version of Zapier or N8n allowing you to build simple automations without the technical know-how. We're watching ChatGPT evolve from a simple chat interface into a complete workspace.

• Sora 2: OpenAI's newest video generation model delivers significantly better realism and gives you way more creative control than before. I break down exactly what this means for your business in the Tool Spotlight section below.

• Perplexity Launches Comet AI Browser: Perplexity just went head-to-head with Google by releasing a browser that doesn't just find information, it actually understands what you're trying to accomplish. lt launched the free version of its Comet browser, which was previously available only as part of its $200-per-month “Max” plan. Since Comet’s release in July, the company claims its waitlist has amassed “millions” of users. Instead of typing queries and clicking through results, the browser anticipates your needs and surfaces relevant answers as you work.

Tools We're Testing

• Arc Browser: I’ve actually been using Arc for over a year and LOVE it! It’s a modern browser with built-in AI features and a sleek sidebar interface that makes juggling multiple tabs and tools much easier than traditional browsers.

• Google Opal: No-code platform for chaining AI models into custom workflows. Create multi-step apps that combine research, content creation, and media generation without coding. This just came out days ago and it is still in the early experimental phase but it looks cool (and actually somewhat similar to ChatGPT’s AgentKit). It’s effectively a chatbot agent builder on the front end with the ability to make technical edits to the workflow on the backend.

Tool Spotlight: OpenAI Sora

This week OpenAI launched Sora 2, and it's not just another AI video model, it's an entirely new social experience 100% of which is populated with AI-generated content (weird, I know). The breakthrough isn't the technical capabilities (other tools like Veo 3 match those), it's how it's packaged. OpenAI created a standalone iOS app (US-only for now) that functions like TikTok exclusively for AI content. The standout feature is Cameo, you record a quick voice and video sample, and Sora creates an AI avatar that you or others (with permission) can insert into any video. 

Sora 2 doesn't just generate clips, it edits them. It cuts between scenes, maintains consistent audio, adds background music and sound effects. The result feels like actual social media content, not raw AI output.

Real business applications for service-based businesses:

• Service explainer videos: Generate multiple versions of "how it works" videos to test which messaging resonates best with prospects

• Social proof content: Create authentic-looking case study videos or influencer style videos showcasing your product or service.

• Training and onboarding: Build consistent instructional videos for new team members or clients without coordinating schedules for filming

• Seasonal marketing campaigns: Quickly produce holiday or event-specific content featuring you or your team without the production overhead

The privacy implications are significant, you're cloning yourself on a social platform. But this represents where AI content creation is heading: complete production workflows that combine video, audio, editing, and effects into finished content. The interface stays user-friendly while the underlying model has finally caught up to competitors, plus unique social features nobody else offers.

Behind the Scenes at Ampra

Last Friday, Ben, Ampra’s lead AI and Automation architect, and I built something that’s now saving us each a few hours every week.

We used to spend up to 15 or 20 minutes writing personalized follow-up emails after every prospect call. Multiply that across dozens of meetings per week, and you're looking at a serious amount of time.

Fast forward, and now we built an automation that watches for when a Google Meet concludes. Then, it automatically drafts a personalized follow-up email based on our meeting transcription and our specific system instruction on who we are, what we offer and how we communicate. It then drops it directly into Gmail as a draft. No copy-pasting. No context-switching. Just open Gmail, review, and send.

Here's what makes it powerful: The system pulls context from the meeting transcript, references specific discussion points, and matches our communication style. But the magic happened in the setup. We spent time upfront training the AI on our actual writing style, our offers, and Ampra's mission and values. This wasn't a plug-and-play solution, we built it intentionally so the output actually sounds like us.

What used to take 20 minutes of drafting from scratch now takes about 2 minutes of review time. We're not just accepting whatever the AI spits out. We're checking for accuracy, making sure the personalization hits right, and verifying every detail reflects the actual conversation. But starting with a solid draft that's already 90% there beats staring at a blank screen after five back-to-back calls.

Want to chat about how we can set this up for you or how we can help automate part of your daily or weekly workflow? Hit reply and share what's on your list of "things I wish didn't take so long."

That's it for the week! Curious what you think about Sora 2 or the automation we built. Hit reply and let me know what you want to see more or less of next week. Have an idea for a spotlight section you want us to dive into? Let’s hear it! 

-Julien