Is it a good thing that the U.S. remains one of the few places in the world where one can order up the sex of your future baby like "selecting the toppings on a pizza"?
Thrilled to share this one with everyone.
In 2009, North Carolina hired analytics firms to investigate whether students were being placed in math classes appropriately. They weren't. More than half of students ready for advanced math were left out.
https://t.co/Djob42hnuY
Don't want to miss that key #GoogleIO session? We've got you covered! The official live stream schedule is live, helping you navigate all the exciting announcements and deep dives. Plan ahead and maximize your learning.
See the full schedule and start planning → https://t.co/q5yPIbajdp
MIT just solved humanity's biggest weakness: In 2025 the average attention span lasts no longer than 12 seconds. But fortunately scientists discovered a 3-step method that fixes this. Here's the protocol to rewire your brain:
New Video Alert – CrewAI: Noob vs Pro – Episode 2
I’m thrilled to announce the release of Episode 2 in my YouTube series: CrewAI: Noob vs Pro! 🎉
This series is all about helping CrewAI developers avoid common mistakes and unlock the full potential of CrewAI.
In this episode, we’re diving into tools—how to create them, integrate them with apps like Trello and Reddit, and use CrewAI’s latest features to automate workflows.
By the end of the video, you’ll learn:
1️⃣ How to create custom tools that seamlessly integrate with external apps.
2️⃣ How to use CrewAI’s newest features (like the Before Kickoff decorator) to make your Crews smarter and faster.
3️⃣ How to build a Crew that automates tasks like researching topics, writing articles, and updating your Trello board.
🔥 Why You’ll Love This Episode:
• Real-World Examples: I break everything down step-by-step to show you exactly how to build efficient tools.
• Practical Applications: See how these tools can automate Trello workflows, pull data from Reddit, and handle complex CrewAI tasks effortlessly.
• Free Source Code: Follow along with the tutorial and grab all the source code—it’s linked in the description!
What do you think about this episode? I’d love to hear your thoughts or see the tools you’re working on! Drop a comment or send me a DM. 😊
Here's a link to the video:
https://t.co/ecAMIRBBcY
Cheers,
Brandon Hancock 🧑💻
My o1-based AI programming agent is now state of the art on SWE-Bench Verified! It resolves 64.6% of issues.
This is the first fully o1-driven agent we know of. And we learned a ton building it.
DeepSeek-R1 (Preview) Results 🔥
We worked with the @deepseek_ai team to evaluate R1 Preview models on LiveCodeBench.
The model performs in the vicinity of o1-Medium providing SOTA reasoning performance! Huge kudos to the team and I'm looking forward to the full release!!
/1
The next billion-dollar tech wave isn't what you think.
While everyone's focused on consumer AI...
Enterprise AI agents are quietly transforming how Fortune 500 companies operate.
But underneath, they all follow a series of similar patterns:
We are happy to announce Curator, an open-source library designed to streamline synthetic data generation!
High-quality synthetic data generation is essential in training and evaluating LLMs/agents/RAG pipelines these days, but tooling around this is still entirely lacking!
So we built Curator, and I think it has easily 10x'ed our productivity in creating post-training datasets!
We are cooking a lot of good features and examples to be released in the future, so you don't want to miss this!
🧵More info below!