What is today’s beat?

• Anthropic’s plans for MCP evolution

• EXO private search to mitigate LLM hallucinations

• Try the EXO wrapper to give a summary of your X account in 2024.

  
  
  Your FREE newsletter
   share
  to show support
  

🎯 RELEASES 🎯

Anthropic
Model Context Protocol Plans for 2025

Synopsis

The Model Context Protocol (MCP), led by Anthropic, is advancing rapidly with a series of planned developments throughout 2025. The protocol is evolving with key updates aimed at expanding accessibility, enhancing agentic workflows, and fostering a more secure, standardised ecosystem. Its impact will be significant in how AI models interact with remote servers, integrate multiple modalities, and support complex workflows in distributed environments.

Core Observations

1. Remote MCP Support: The protocol is prioritising enabling remote connections to MCP servers, incorporating features like OAuth 2.0-based authentication, service discovery, and stateless operations for serverless environments.

2. Distribution & Discovery: Efforts include standardised packaging for servers, simplified installation tools, sand-boxing for better security, and the creation of a server registry for easier server discovery.

3. Agent Support Expansion: Future developments will support hierarchical agent systems, interactive workflows with user permissions, and real-time updates for long-running agent tasks.

4. Broader Ecosystem Collaboration: MCP is looking to foster community-led standards and expand its capabilities beyond text, incorporating audio, video, and more diverse formats into the protocol. View their Github Discussions for active topics.

Broader Context

The roadmap highlights the MCP's role in enabling more flexible, scalable, and secure interactions between AI models and servers. The focus on remote connectivity, agent-based workflows, and multi-modal support positions MCP as a central component in future AI ecosystems. By allowing a wide range of contributors to shape its development, MCP aims to become a universal standard that can adapt to various applications across industries, facilitating smoother integration and innovation within the rapidly growing field of artificial intelligence.

Read about their plans here.
Read their MCP November blog here

EXO
Private Search: A Step Toward Secure, Real-Time AI

Synopsis

Exo Labs is making strides in private, real-time AI interactions through their development of DeepSeek V3 and associated technologies. The company focuses on enhancing privacy while delivering accurate, up-to-date AI model outputs. By leveraging homomorphic encryption and optimised clustering for search, Exo aims to address the challenge of maintaining privacy while accessing real-time information.

Core Observations

1. Real-Time Search Integration: Exo integrates real-time search to avoid common LLM problem of hallucination. It augments the model’s output with relevant search engines results.

2. Homomorphic Encryption: performs operations on encrypted data, allowing the server to compute results without accessing the raw query, thus ensuring privacy during search queries.

3. Optimized Search Efficiency: their approach partitions large datasets into smaller clusters, reducing the number of comparisons and speeding up the retrieval process.

Broader Context

Exo Labs' positions itself in a significant role in advancing AI by addressing some of the key concerns of modern AI applications - data privacy and model hallucinations. They provide tools that allow secure, local processing of AI model outputs with real-time data integration. Exo positions itself as a key player in the shift toward privacy-conscious, decentralised AI applications. This has far-reaching implications for industries where data security and up-to-date information are critical, with special interests in heavily regulated industries like healthcare, finance, and personal AI systems.

Read here

Trending

• Sonus-1, a A New Era of LLMs released

• Deloitte, 16th annual Tech Trends report - 2025 at a Glance

• Exo (LLM search), view their 12 days of release 

We are 100% free!

And with your support, we create more FREE content!

Please share us with a friend!

⚙️ BUILDERS BYTES ⚙️

What will you learn today?

This tutorial walks you through creating a real-time push-to-talk audio application using OpenAI's GPT with asyncio and the Textual framework.

Key Takeaways

1. Terminal Interface: quick prototyping with interactive terminal app.

2. Real-time with AsyncIO: audio streaming with OpenAI's API.

3. Session-based connection: OpenAI's API for live audio communication per use (uuid session).

⭐️ ⭐️⭐️⭐️⭐️
Like these tutorials?
👉️ Star out repo to show support
⭐️⭐️⭐️⭐️⭐️⭐️

Do you have a product in AI and would like to contribute?
👉️ email us: [email protected] 

Is there something you’d like to see in this section?
👉️ share your feedback

Trending

• Minima: On-premises conversational RAG with configurable containers

• RWKV, a Linux Foundation AI project

• IBM Granite 3.1 Models with release (recently on HuggingFace)

Research

• OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis

• A3: Android Agent Arena for Mobile GUI Agents

• CodeElo: Benchmarking Competition-level Code Generation of LLMs with Human-comparable Elo Ratings

🤩 COMMUNITY 🤩

TOOLS

• Replicate: Run AI with an API - Run and fine-tune open-source models. Deploy custom models at scale

• Asterai's serverless platform allows you to easily and rapidly build, integrate, deploy, and scale AI applications

• Track & Compare LLM API Performance Metrics - a leaderboard trackrecord of providers

• A EXO wrapper, gives a 2024 review of your X.com account

THANK YOU

Found something cool?
Want something different?

Our Mission at AlphaWise

AlphaWise strives to cultivate a vibrant and informed community of AI enthusiasts, developers, and researchers. Our goal is to share valuable insights into AI, academic research, and software that brings it to life. We focus on bringing you the most relevant content, from groundbreaking research and technical articles to expert opinions to curated community resources. 

Looking to connect with us?

We actively seek to get involved in community with events, talks, and activities. Email us at [email protected] 

Looking to promote your company, product, service, or event?