🦌 ByteDance launches DeerFlow 2.0

Good Morning! ByteDance just open-sourced a massive new AI framework called DeerFlow 2.0, and it completely blew up on GitHub. Plus, I’ll show you how to find the cheapest gas using AI in Google Maps in seconds.

Plus, in today’s AI newsletter:

• ByteDance Drops DeerFlow 2.0 "SuperAgent" Harness

• Tencent Drops OpenClaw Agent into WeChat

• Google’s New Blueprint for Measuring AGI

• How to Find the Cheapest Gas Using AI in Google Maps

• 4 new AI tools worth trying

AI TOOLS

🦌 ByteDance Drops DeerFlow 2.0 "SuperAgent" Harness

China's ByteDance just released DeerFlow 2.0, a highly capable open-source framework designed to act as a runtime environment for orchestrating autonomous AI sub-agents. It recently hit #1 on GitHub Trending and is built to handle complex, long-horizon tasks that take minutes or even hours to finish.

• Operates in isolated Docker sandboxes, giving the AI its own literal "computer" with a persistent filesystem and bash terminal to safely execute code

• Uses "Progressive Skill Loading," injecting specific capabilities into the context window only when needed to keep token usage lean and incredibly efficient

• The lead agent automatically decomposes complex prompts, spawning scoped sub-agents that run in parallel to research, code, and synthesize deliverables

• Features persistent, cross-session long-term memory so the agent actually learns your preferences and workflows over time

Most agent frameworks are either thin wrappers around API calls or highly opinionated coding assistants. DeerFlow 2.0 represents a massive leap by giving AI an actual secure environment (with a filesystem, memory, and sandboxing) to live and work in. By open-sourcing a system capable of managing multi-hour, complex workflows, ByteDance is pushing us straight into the era of managing fully autonomous digital employees rather than just prompting chatbots.

AGENTIC AI

🐉 Tencent Drops OpenClaw Agent into WeChat

Tencent just launched ClawBot, a new tool that integrates the viral open-source AI agent, OpenClaw, directly into its massive WeChat messaging platform. As tech giants rush to capitalize on the agentic AI boom, Tencent is leveraging its 1 billion+ monthly active users to take the lead.

• ClawBot appears as a standard contact within WeChat, allowing users to connect with and control OpenClaw directly through the messaging interface

• Users can prompt the agent to autonomously execute tasks on their behalf, such as transferring files and sending emails

• This follows Tencent's recent launch of its broader AI agent suite: QClaw for individuals, Lighthouse for developers, and WorkBuddy for enterprises

• The Chinese AI market is rapidly accelerating: Alibaba just launched "Wukong," a platform that coordinates multiple agents for complex business tasks, while Baidu rolled out its own series of OpenClaw-built agents for desktop, cloud, and smart-home devices

WeChat is the digital backbone of China. By baking an autonomous agent directly into an app used by over a billion people daily, Tencent isn't just launching a new feature, they are normalizing agentic AI for the general public at an unprecedented scale. With Alibaba and Baidu aggressively pushing their own agent platforms, the race to dominate the everyday autonomous assistant market is officially in hyperdrive.

AI RESEARCH

🧠 Google’s New Blueprint for Measuring AGI

Google's latest paper proposes a completely new way to track our progress toward AGI. Rather than treating AGI as a single finish line, the researchers have developed a framework to map out the exact cognitive capabilities of frontier models.

• Instead of just asking "is this AGI?", the paper introduces a framework that tests AI models on a wide variety of held-out tasks.

• The methodology systematically compares model performance against strict human baselines to establish a grounded, realistic benchmark.

• It ultimately builds a detailed profile highlighting exactly where the AI matches human cognitive abilities and where it still falls short.

As frontier models get more advanced, relying on basic benchmarks or moving goalposts isn't enough. By standardizing how we evaluate AGI against actual human cognition, we get a much clearer, practical map of where AI is truly superhuman and where the gaps remain.

HOW TO AI

🗂️ How to Find the Cheapest Gas Using AI in Google Maps

In this tutorial, you will learn how to instantly locate the most affordable gas stations and optimize your commute using the new AI-powered Ask Maps feature, saving you from doing manual research and paying premium prices.

🧰 Who is This For

• Small business owners looking to cut down on daily transportation or fleet costs to improve profit margins

• Daily commuters who want to stop overpaying for gas without going out of their way

• Content creators looking for highly practical, everyday AI use cases to share

• Anyone who wants to leverage conversational AI for real-world, localized recommendations

STEP 1: Access the AI Conversational Experience

Open the Google Maps app on your phone to get started. Instead of just typing a generic location into the standard search bar, look for the newly updated "Ask Maps" conversational interface. This upgrade uses advanced AI to process complex, natural language questions rather than relying on simple keyword matching.

STEP 2: Ask for the Best Local Prices

Tap into the Ask Maps input field and simply type or say, "Where's the cheapest gas near me?" The AI will immediately analyze real-time pricing data across your surrounding area, bypassing the need for you to click into individual station profiles to check their current rates.

STEP 3: Compare and Choose Your Stop

Google Maps will generate a curated list of the best options right on your screen. Take a moment to review the AI's recommendations, weighing the current price against the distance from your location and overall convenience. This allows you to quickly pick the optimal station that makes the most sense for your wallet and schedule.

STEP 4: Optimize Active Routes

If you are already navigating to a destination, utilize the "Search along route" feature instead. You can instruct Maps to locate the cheapest gas stations within a specified distance from your current path. This gives you affordable options that might be slightly off your main route but are still convenient enough to stop at without making a separate, time-consuming trip.

Sources describe efforts to build and use internal AI tools at Meta, including Mark Zuckerberg building a CEO agent to help him do his job better.

Elon Musk announces Terafab, an Austin-based project run by Tesla and SpaceX to manufacture robotics, AI, and space data center chips for Tesla, xAI, and SpaceX.

Amazon begins three-hour deliveries in ~2K US cities and towns and one-hour deliveries in hundreds of those areas, after 2025 pilots; 90K+ products are eligible.

🎨 MAI-Image-2: Microsoft’s improved image generator

🚀 Google AI Studio: Google’s upgraded all-in-one AI builder

⚙️ OpenAI Codex: OpenAI’s coding assistant, now with automations and customizable themes

🎥 Replit Animation: Turn text prompts into animated videos

THAT’S IT FOR TODAY

Thanks for making it to the end! I put my heart into every email I send, I hope you are enjoying it. Let me know your thoughts so I can make the next one even better!

See you tomorrow :)

- Dr. Alvaro Cintas