Future Forward - 78th Edition - Last Week in AI - A Primer on DSPy.

Welcome to the 78th Edition of Future Forward - the Emerging Tech & AI Newsletter!

This newsletter aims to help you stay up-to-date on the latest trends in emerging technologies and AI. Subscribe to the newsletter today and never miss a beat!

Subscribe to the newsletter here.

Each edition covers top AI news from last week and an AI-related topic - Primers/Tutorials/ How AI is being used.

Here's what you can expect in this issue of the Emerging Tech & AI Newsletter:

• A summary of the top AI news from the past week.
• A Primer on DSPy.

AI News from Last Week

The field of AI is experiencing rapid and continuous progress in various areas. Some of the notable advancements and trends from the last week include:

Big Tech in AI:

1. Github announced agent mode for GitHub Copilot.
2. Google is bringing SynthID to Reimagine in Magic Editor.
3. Gemini 2.0 was made available to everyone.
4. Nvidia and researchers released paper on ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills.
5. Google updated its ethical AI guidelines.
6. Amazon's AI revamp of Alexa assistant nears unveiling.
7. Meta release Frontier AI framework.
8. Apple Began Mass Production of AI-Optimized M5 Chip for Next-Gen Devices.

Funding & VC Landscape:

1. LogicStar raised $3M pre-seed.
2. Prior Labs raised €9M Pre-Seed.
3. Neofin raised $7M Seed funding.
4. Ivo raised $16M Series A.
5. File AI raised $14M in Series A.
6. Salience Labs raised $30M.
7. Lorikeet raised $9M.
8. & AI raised $6.5M Seed Funding.
9. Visarun.ai raised $700K Pre-Seed.
10. GetWhys raised $2.75M Seed Funding.
11. ThreatMate raised $3.2M Seed Funding.
12. TrueFoundry raised $19M Series A.
13. Cognida.ai raised $15M Series A.
14. INFI USA raised $12M Series A.

Other AI news:

1. Mistral unveiled Android and iOS app for its 'le chat' assistant.
2. Lyft partnered with Anthropic to deploy Claude for customer service.
3. Two.ai released SUTRA-R0, its model for indic languages.
4. User-friendly system can help developers build more efficient simulations and AI models.
5. Hugging Face released DeepResearch, open source equivalent of Open AI's similar tool.
6. Adobe Acrobat’s AI chatbot can now decipher contract jargon.
7. OpenAI and SoftBank Group Partner to Develop and Market Advanced Enterprise AI.
8. Anthropic released Constitutional Classifiers.
9. Snap Developed Groundbreaking AI Text-to-Image Model.
10. ByteDance released OmniHuman-1.
11. Pika.ai released pickadditions.
12. Replit launched the first software creation Agent on iOS and Android.
13. Hugging Face launched AI App Store - Spaces.

Liked the news summary? Subscribe to the newsletter to keep getting updates every week.

A Primer on DSPy

DSPy is an open-source Python framework that allows developers to build language model applications using modular and declarative programming instead of relying on one-off prompting techniques. It was created by Stanford University. Quoting from official github repo:

"DSPy is the framework for programming—rather than prompting—language models. It allows you to iterate fast on building modular AI systems and offers algorithms for optimizing their prompts and weights, whether you're building simple classifiers, sophisticated RAG pipelines, or Agent loops.

DSPy stands for Declarative Self-improving Python. Instead of brittle prompts, you write compositional Python code and use DSPy to teach your LM to deliver high-quality outputs."

Key Features:

• Declarative programming: Define the task and metrics for success, and DSPy optimizes the models behavior.
• Self-improving prompts: Automatically improve prompts over time using feedback and evaluation.
• Modular architecture: Mix and match pre-built modules for different NLP tasks.

Research Behind DSPy:

DSPy is based on the idea of compiling declarative language model calls into self-improving pipelines. Instead of manually crafting prompts, developers can use DSPy to program the AI models directly. This makes AI apps more reliable and easier to scale. DSPy separates the apps logic from the text it uses, allowing developers to focus on what they want the AI to do.

DSPy optimizes prompts using:

• In-context learning
• Automatic few-shot example generation

This means that the framework continuously refines the prompts to improve the models performance. DSPy can also fine-tune smaller models for tasks requiring more specific tuning.

Benefits of Using DSPy:

• Improved reliability
• Simplified development
• Adaptability
• Scalability

DSPy can be applied to a wide range of use cases, including question answering, text summarization, code generation, and custom NLP tasks.

In summary, DSPy is a powerful tool that simplifies the development of language model applications. Its declarative approach, self-improving prompts, and modular architecture make it a valuable asset for any developer working with AI.

Research Papers:

Here's link to related research papers

[Jul'24] Fine-Tuning and Prompt Optimization: Two Great Steps that Work Better Together

[Jun'24] Optimizing Instructions and Demonstrations for Multi-Stage Language Model Programs

[Jun'24] Prompts as Auto-Optimized Training Hyperparameters

[Feb'24] Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models

[Jan'24] In-Context Learning for Extreme Multi-Label Classification

[Dec'23] DSPy Assertions: Computational Constraints for Self-Refining Language Model Pipelines

[Oct'23] DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines

[Dec'22] Demonstrate-Search-Predict: Composing Retrieval & Language Models for Knowledge-Intensive NLP

Disclosure: The content in "A Primer on DSPy" section was written with the help of Google Gemini using Flash 2.0 model. Please write to us in case of any gaps.

Thanks for reading. See you next week!

Let's explore the future of technology together!

Your Turn:

Did you like the content? Share with your network. Subscribe now to get updates directly in your mailbox.