• AI Horizons
  • Posts
  • ChatGPT Wants To Be Your Assistant, AI Creates New Materials from Scratch, and Healthcare AI Hits Unicorn Status

ChatGPT Wants To Be Your Assistant, AI Creates New Materials from Scratch, and Healthcare AI Hits Unicorn Status

PLUS: Open-Source AI Breakthroughs, Trending Tools, and Learn to Code for AI

In partnership with

We’re back with another issue of the latest breakthroughs, insights, and tools to keep you ahead of the curve. Whether you’re an AI engineer, business leader, developer, or creator, we’ve got everything you need to stay informed and inspired.

Here's what we're covering today:

  • ChatGPT’s New "Tasks" Feature

  • MatterGen’s AI Breakthrough in Materials Science

  • Hippocratic AI’s Rapid Healthcare Growth

  • Advanced AI for Just $450

  • Tools & Resources to Level Up Your AI

FEATURED INSIGHT💡

ChatGPT Introduces "Tasks": Your Personal AI Assistant

OpenAI has unveiled Tasks, a beta feature that allows users to schedule reminders and automate recurring actions seamlessly within ChatGPT.

How it works:

  • Schedule personalized actions like daily weather reports, weekly news updates, or reminders for passport renewals.

  • Manage tasks via chat or through the "Tasks" section in the profile menu, available on web, desktop, and mobile platforms.

  • ChatGPT will also suggest relevant tasks based on your previous interactions, adding a proactive layer to its capabilities.

Why it matters: This feature signals a significant shift for ChatGPT from a reactive chatbot to a proactive assistant. Tasks can streamline productivity, making it easier for users to integrate AI into their workflows. For now, it’s exclusive to Plus, Pro, and Teams users but will roll out more broadly soon.

ON THE HORIZON 🌅

MatterGen: Revolutionizing Material Discovery with Generative AI

Microsoft Research’s MatterGen is pioneering a new era in materials science by using generative AI to design novel crystalline materials tailored to specific properties.

Why it’s better than previous approaches: Traditional methods rely on labor-intensive trial-and-error experimentation or computational screening confined to known materials databases. In contrast, MatterGen uses a 3D diffusion-based generative model to directly create new materials, overcoming these limitations.

Key Advantages:

  • Efficiency: It generates stable, unique, and novel materials faster than screening-based methods.

  • Flexibility: Fine-tunes materials for specific properties like mechanical strength, electronic conductivity, or sustainability.

  • Innovation: Unlocks the potential to discover materials beyond current databases, significantly broadening the design space.

Real-world impact: MatterGen’s predictions have already led to the synthesis of new materials like TaCr₂O₆, with experimental results closely matching AI-generated targets. This breakthrough could transform industries from energy storage to quantum computing, making material innovation faster, cheaper, and more sustainable.

(For the technically inclined, they open-sourced their code!)

LATEST IMPORTANT NEWS 📰

Hippocratic AI: Transforming Healthcare with Generative AI

Hippocratic AI has raised $278M in Series B funding, achieving unicorn status as it builds a large language model (LLM) for healthcare applications.

What’s new:

  • Launched an AI agent app store, allowing clinicians to design and deploy AI agents tailored to patient care needs.

  • These agents assist with tasks like chronic care management, post-discharge follow-ups, and disaster response.

  • Ensures safety through a rigorous multi-phase testing process, involving over 6,000 nurses and 300 doctors.

Why it matters: With a growing healthcare staffing shortage, generative AI tools like Hippocratic AI provide scalable solutions to improve patient care while reducing the burden on human clinicians.

FOR THE TECHNICALLY INCLINED 🛠️

Sky-T1: Open-Source AI for Advanced Reasoning

UC Berkeley’s NovaSky team has developed Sky-T1-32B-Preview, a reasoning model that rivals OpenAI’s o1-preview but costs just $450 to train.

Key Features:

  • Fine-tuned on Alibaba’s Qwen2.5-32-Instruct, the model excels in coding and mathematical reasoning benchmarks.

  • Entirely open-source, including training data, code, and model weights, enabling the community to replicate and innovate.

How they did it:

  • Curated high-quality training data using advanced rejection sampling to improve performance across math and coding domains.

  • Trained the model on 8 H100 GPUs in just 19 hours using efficient techniques like DeepSpeed Zero-3.

Why it matters: Sky-T1 demonstrates that cutting-edge reasoning capabilities are no longer exclusive to large-scale AI labs. This open-source milestone empowers smaller teams to contribute to AI advancements, fostering a more collaborative and inclusive research landscape.

An entirely new way to present ideas

Gamma’s AI creates beautiful presentations, websites, and more. No design or coding skills required. Try it free today.

AI TOOL OF THE DAY 🛠️🚀

ReplyPilot: Smarter Social Media Interactions. ReplyPilot uses AI to craft personalized, on-brand responses across social media platforms, saving you time while maintaining authenticity. With tone learning and multilingual support, this tool ensures every reply aligns perfectly with your brand’s voice.

That's all for now!

We'll catch you in the next one.

Cheers,

The AI Horizons Team

P.S. If you missed our last issue, no worries, you can check out all previous issues here!

P.P.S We value your thoughts, feedback, and questions - feel free to respond directly to this email!

... and if you enjoyed this email, be sure to thank the person who forwarded it to you and subscribe to get your own copy!

What did you think about today's email?

Login or Subscribe to participate in polls.