Beat China or Bust

Trump's AI action plan ditches regulation for growth, ChatGPT Plus gets meeting superpowers, and Google's Gemini can now slice images with words...

In this edition we’ll be covering…

• Trump's massive AI deregulation strategy to beat China

• How to use ChatGPT's new meeting recording feature on Mac

• A breakdown of Google's conversational image segmentation with Gemini 2.5

• 5 trending AI signals from the week

• 3 AI tools to boost your productivity game

• And much more…

The Latest in AI

Trump Trades AI Safety For Speed In China Race

The Trump administration just dropped its much-anticipated AI Action Plan, and it's clear they're taking the "move fast and break things" approach to AI development.

With over 90 federal actions outlined, the plan prioritizes American dominance over safety guardrails...

Key elements driving this AI acceleration:

• The plan eliminates references to diversity, equity, inclusion (DEI), climate requirements, and misinformation from federal AI guidelines, allowing companies to develop with fewer constraints and biases.

• New categorical exclusions and streamlined permitting processes will accelerate data center construction on federal lands, including national parks and military bases.

• Federal agencies will only contract with AI model developers who ensure their systems are "objective and free from top-down ideological bias."

So What?

While this aggressive approach may accelerate American AI innovation and help compete with China's rapid progress, it raises serious questions about safety oversight and potential societal risks.

The plan essentially lets the tech industry self-regulate while using taxpayer dollars to build their infrastructure. Trump's betting that speed trumps safety in the AI arms race – time will tell if that gamble pays off or backfires spectacularly.

Tool Spotlight

Record and Summarize Meetings with ChatGPT

ChatGPT Plus subscribers can now turn their Mac into a powerful meeting assistant using the new Record mode.

Instead of paying for a separate AI meeting assistant, you can now just use ChatGPT!

Here's how to set up meeting recording with ChatGPT:

1. Open the ChatGPT Mac app (this feature is currently available on the Desktop app for Mac only, but Windows support is coming soon).

2. Click the Record button at the bottom of the screen when you're ready to start capturing audio.

3. Grant necessary permissions the first time you use the feature – ChatGPT needs microphone access to record conversations.

4. Click Stop when finished, then hit Send to process the audio through ChatGPT's transcription engine.

🔥 Pro Tip: Switch to the o3-mini model for this specific chat – it's optimized for analytical tasks, delivering precise results for meeting summaries and action items.

Once processed, ChatGPT will provide a comprehensive text summary with key points, action items, and suggested follow-up questions. You can then ask specific questions about the conversation or request different formats like email summaries or task lists.

Get Your Hands Dirty!

Activate Conversational Image Segmentation with Gemini 2.5

Google just unleashed something incredible - Gemini 2.5's ability to understand and segment images using natural language.

Forget pointing and clicking – now you can literally tell AI what parts of an image you want to isolate.

Here's how developers can implement this feature:

1. Access Google AI Studio via the interactive Spatial Understanding demo or fire up the Spatial Understanding colab notebook for a Python environment.

2. Use the gemini-2.5-flash model for optimal performance and set thinkingBudget to 0 for best results.

3. Upload an image or choose a default one. Structure your prompts properly:

Give the segmentation masks for the objects. 

Output a JSON list of segmentation masks where each entry contains the 2D bounding box in the key "box_2d", the segmentation mask in key "mask", and the text label in the key "label". 

Use descriptive labels.

The magic lies in the types of queries you can make. Instead of simple labels like "car," you can ask for complex requests like "the person holding the umbrella," "food that is vegetarian," or even abstract concepts like "areas that should be cleaned up."

This opens massive possibilities for cases in content creation safety monitoring, and insurance damage assessment where traditional object detection might fall short.

Quick Bites

Stay updated with our favorite highlights, dive in for a full flavor of the coverage!

Amazon acquired Bee, the AI wearable startup that records everything you say to create reminders and to-do lists, signaling Amazon's push into wearable AI beyond Echo devices.

OpenAI CEO Sam Altman warned financial executives of an impending AI fraud crisis, specifically targeting voice authentication systems that could enable massive money transfers.

YouTube Shorts launched image-to-video AI tools powered by Veo 2, letting creators turn static photos into six-second videos with "subtle movements" or "I'm feeling lucky" prompts.

Large language models have transformed the way we build software systems. In Chroma’s latest research report, Kelly Hong shares her findings on “Context Rot”, how large language model performance degrades with more and more input tokens.

Forget vendor hype: Is AI actually boosting developer productivity, or just shifting bottlenecks? Stop guessing. This study at Stanford cuts through the noise, analyzing real-world productivity data from nearly 100,000 developers across hundreds of companies.

Trending Tools

🔍 Devi AI - Helps you find organic leads by monitoring Facebook private and public groups, LinkedIn, Twitter, Bluesky, Reddit, and Telegram posts. Save 3 hours a day and grow your business. Get started here.

🌐 Google Agentspace - enables AI agents to securely access your most-used applications to quickly find what you need—making you faster, smarter, and more impactful at your job.

🗣️ TheySaid.io - 10x your insights with AI surveys and interviews.

The Neural Network

Turns out the big AI labs aren't just stealing each other's talent, they're actually working together behind the scenes.

While we're all watching the public drama of OpenAI vs Google vs Anthropic, the real story might be one big happy AI family sharing notes…

Until we Type Again…

Thank you for reading yet another edition of Digestible AI. Be sure to give us a follow on X, Instagram, and LinkedIn too!

If you have any suggestions or specific feedback, simply reply to this email. Additionally, if you found this insightful, don't hesitate to engage with us on our socials and forward this over to your friends!