OpenAI's Developer Surprise

ChatGPT drops new models and Google is trying to talk to dolphins…

In this edition we’ll be covering…

• The new GPT-4.1 model

• A breakdown on Google’s DolphinGemma

• A tutorial on how to use Brave Search’s MCP server directly in Cursor

• 5 more pieces of AI news

• 3 new AI tools

• And much more…

The Latest in AI

OpenAI’s Naming Still Sucks, But GPT-4.1 Doesn’t

OpenAI just dropped GPT-4.1, and it’s exclusively for devs. For now, it’s only available in the API, but we’ll track to see how that changes.

The new model packs a million-token context window (yes, 1,000,000, new ground for OpenAI), massive improvements in reasoning and instruction following, and beats GPT-4o “in just about every dimension,” according to the company.

What’s New?

• Longer memory: GPT-4.1 can handle prompts up to 1 million tokens, far beyond GPT-4o’s 128K limit. It’s built to stay focused across those long stretches and zero in on key details along the context.

• Better at code & reasoning: On SWE-Bench, GPT-4.1 completed 54.6% of tasks—a 21% boost over GPT-4o and a clear sign it’s becoming a true coding companion.

• Cheaper & faster: It’s 26% more cost-efficient than GPT-4o, which feels perfectly timed after DeepSeek started stirring the pot on price performance.

• Mini & Nano models: OpenAI’s also shipping GPT-4.1 Mini and Nano, lighter, faster versions designed for devs who want top-tier performance without melting their GPU or wallet.

So What?

Why are we going backwards with naming conventions? Don’t ask, OpenAI was never good at naming its products.

OpenAI is phasing out GPT-4 and GPT-4.5 preview to make room for this upgrade, while GPT-5 is reportedly delayed as the team works to integrate more complex features.

If you’re building apps, writing code, or just want an AI that doesn’t forget what you said 5 seconds ago, GPT-4.1 just might be the new meta.

This reveal also confirms that OpenAI will soon launch EVEN MORE models this week, so stay tuned for the fun…

Together with Artisan

Try Artisan’s All-in-one Outbound Sales Platform & AI BDR

Ava automates your entire outbound demand generation so you can get leads delivered to your inbox on autopilot. She operates within the Artisan platform, which consolidates every tool you need for outbound:

• 300M+ High-Quality B2B Prospects, including E-Commerce and Local Business Leads

• Automated Lead Enrichment With 10+ Data Sources

• Full Email Deliverability Management

• Multi-Channel Outreach Across Email & LinkedIn

• Human-Level Personalization

Book a demo to see what Ava can do.

Industry Intel

Google DolphinGemma Takes Sea to CPU

What if your phone could help you talk to dolphins? Well that’s not a Black Mirror plot, it’s Google’s real project.

On National Dolphin Day, Google unveiled DolphinGemma, an AI model designed to decode and generate dolphin vocalizations. Built in collaboration with Georgia Tech and the Wild Dolphin Project (WDP), DolphinGemma is trained on decades of underwater audio and video from Atlantic spotted dolphins off the coast of the Bahamas.

Wait, what?

• Since 1985, WDP has tracked the social lives and sounds of individual dolphins, linking signature whistles and buzzes to real behaviors like hunting, fighting, or calling for their calves.

• DolphinGemma processes natural dolphin sequences and predicts what sounds come next—kind of like how ChatGPT predicts your next word, but with clicks and whistles instead of text.

• The model is small enough (~400M parameters) to run on a Google Pixel 9.

• Google plans to release DolphinGemma as an open model this summer, and who knows what that can potentially unlock for interspecies communication!

So What?

This is a first step toward genuine interspecies communication, and it’s powered by AI. With DolphinGemma, we’re inching closer to understanding one of Earth’s most intelligent creatures on their terms, not ours.

Next time someone says AI is making us more disconnected from nature… tell them it’s helping us talk to dolphins.

Tool Spotlight

How to Integrate Brave Search with Cursor Using MCP

So MCP is sweeping the web, and one of the clients you can use it with is Cursor.

Integrating Brave Search via MCP allows you to perform real-time web searches directly within your coding environment.

This is particularly useful for tasks like fetching documentation, exploring code examples, or staying updated with the latest development trends without leaving Cursor.

Here’s how you can do it:

1. Open up Cursor (or download/install it here). You’ll also need a Brave Search API Key.

2. Go to Settings → Cursor Settings.

3. Jump down to MCP on the left and hit Add new global MCP server.

4. This will open up mcp.json, go ahead and paste the npx command:

{
  "mcpServers": {
    "brave-search": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-brave-search"
      ],
      "env": {
        "BRAVE_API_KEY": "YOUR_API_KEY_HERE"
      }
    }
  }
}

5. Now when you call a tool that can use Brave, Cursor will search the web in real time!

Quick Bites

Stay updated with our favorite highlights, dive in for a full flavor of the coverage!

NVIDIA is working with its manufacturing partners to design and build factories that, for the first time, will produce NVIDIA AI supercomputers entirely in the U.S.

In this talk from the AI Engineer YouTube channel, Douwe Kiela shares lessons learned from deploying enterprise RAG systems at scale and how to design a system robust enough for the Fortune 500.

Johns Hopkins students use AI to help develop better baseball bats.

Netflix is starting to test search that’s powered by OpenAI. The new search engine will let users “look for shows using far more specific terms, including the subscriber’s mood, for example.

Safe Superintelligence (SSI), the AI startup led by OpenAI’s co-founder and former chief scientist Ilya Sutskever, has raised an additional $2 billion in funding at a $32 billion valuation.

Trending Tools

🤖 MCPify - Create custom MCP servers without code.

❤️ Lovable - AI-powered tool for building software products without writing code.

💸 Jupid - AI accounting solution that helps you save on taxes.

The Neural Network

According to our last poll, you guys are ready to VIBE CODE. I’d highly recommend checking out Lovable above too and add it to your stack!

I thought it would be fitting to throw this in too…

Until we Type Again…

Thank you for reading yet another edition of Digestible AI. Be sure to give us a follow on X, Instagram, and LinkedIn too!

If you have any suggestions or specific feedback, simply reply to this email. Additionally, if you found this insightful, don't hesitate to engage with us on our socials and forward this over to your friends!