Sam's Master Plan Revealed

We’re finally getting our public preview of GPT-5…

In this edition we’ll be covering…

• OpenAI’s revamped roadmap for its next frontier models

• Workday’s big bet on AI agents

• A guide on how to build your own Web Browsing Operator

• 5 trending AI signals

• 3 new AI tools

• And much more…

The Latest in AI

OpenAI Spills its Plans for GPT-5

Sam Altman just dropped OpenAI’s latest roadmap, and big changes are coming. Ironic for OpenAI here, but it looks like they’re going to be more transparent to the community. They just might have nailed their naming conventions in the process…

The company is working to simplify its increasingly complex lineup and roll out two major updates: GPT-4.5 and GPT-5.

What’s coming?

• GPT-4.5 (what was internally called Orion) is OpenAI’s final non-chain-of-thought model, set to launch in weeks/months according to Altman.

• GPT-5 is meant to be unified system combining the best of OpenAI’s GPT-series and o-series models, including o3. No more standalone o3 releases.

• Free users will get unlimited chat access at a “standard intelligence” level, while Plus and Pro subscribers will gain progressively higher intelligence tiers.

• OpenAI wants to ditch the confusing model picker and return to a seamless, “magic” experience.

So What?

OpenAI is pivoting toward a single, more intuitive AI system—one that adapts in real time, reasoning when necessary and streamlining tasks without manual selection. The move reflects a broader industry push toward AI agents that handle complex workflows effortlessly.

And while Elon's throwing billions at the door, Sam's busy making AI that doesn't need a PhD to operate (I secretly think they’re friends).

Together with 1440

The Daily Newsletter for Intellectually Curious Readers

If you're frustrated by one-sided reporting, our 5-minute newsletter is the missing piece. We sift through 100+ sources to bring you comprehensive, unbiased news—free from political agendas. Stay informed with factual coverage on the topics that matter.

Join for free today!

Innovation Showcase

The Office Gets an Upgrade with Digital Workers

AI agents aren’t just for automating simple tasks anymore—Workday CEO Carl Eschenbach is betting big on AI-powered “digital employees” that learn, evolve, and take on entire roles within companies.

Move over human resources, Workday's getting into the digital employee business, and they're creating co-workers who never ask for coffee breaks or complain about Friday meetings…

Key Announcements:

• The company announced the “Agent Hub” on Tuesday, a centralized platform for businesses to manage AI agents, define roles, track impact, and budget costs—similar to onboarding human employees.

• Workday is launching specialized agents for payroll, financial auditing, and contract management.

• CEO Carl Eschenbach envisions a future where AI agents and human employees “peacefully coexist,” with AI handling routine work while humans shift to higher-value tasks.

So What?

While Microsoft, Salesforce, and Oracle are racing to automate everything with a pulse, Workday's jumping in with both circuits. They insist AI won't replace humans but will "re-skill" them.

However, this announcement comes just a week after Workday laid off 1,750 employees (8.5% of its workforce)—raising inevitable questions about AI’s role in workforce restructuring.

Tool Spotlight

AI Goes Shopping with Browser Use

Building AI agents that can use the web autonomously is one of the most exciting frontiers right now.

With that, I'm on the hunt for the best open-source OpenAI Operator alternative, and I recently came across Browser Use.

So naturally, I had to put it to the test.

This tool is completely free to use, and I decided to give it a live test by searching for some good laptops on Amazon. 👀

Here's how you can set up your own AI web browser:

1. Install Browser Use and playwright on your machine:

pip install browser-use

playwright install

2. Add the relevant API keys to call your favorite model to a .env file.

3. Spin up your agent. Change the task and model as needed!

from langchain_openai import ChatOpenAI
from browser_use import Agent
import asyncio
from dotenv import load_dotenv
load_dotenv()

async def main():
    agent = Agent(
        task="Go to Reddit, search for 'browser-use', click on the first post and return the first comment.",
        llm=ChatOpenAI(model="gpt-4o"),
    )
    result = await agent.run()
    print(result)

asyncio.run(main())

Quick Bites

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

Apple is partnering with Alibaba to roll out artificial intelligence features for iPhone users in China.

Dwarkesh Patel interviews Jeff Dean and Noam Shazeer about their 25 years at Google, going from PageRank to MapReduce to the Transformer to MoEs to AlphaChip – and soon to ASI.

Google is bringing its upgraded note-taking app to its One AI Premium plan. That means subscribers will now gain access to NotebookLM Plus at no added cost, which offers higher usage limits and premium features.

China’s Baidu plans to release the next generation of its artificial intelligence model in the second half of this year, according to a source familiar with the matter, as newer players such as DeepSeek disrupt the segment.

GitHub Copilot’s new agent mode is capable of iterating on its own code, recognizing errors, and fixing them automatically. It’s now available in preview.

Trending Tools

🧞 PixelGenie - Create stunning AI-generated portraits of your beloved pet in any scene.

🤖 n8n - The world's most popular workflow automation platform for technical teams.

✂️ Xtract - a powerful AI-driven Chrome extension that revolutionizes how you interact with web content, powered by Google's Gemini.

The Neural Network

So NVIDIA quietly dropped this new model called ASAP that teaches bots how to act like humans through simulations

Let’s just say we're watching the birth of robot athletics, folks. Soon we might have humanoid robots crossing defenders and hitting fadeaway jumpers. 👀

Just don't expect them to sign autographs – they haven't quite mastered that yet…

Until we Type Again…

Thank you for reading yet another edition of Digestible AI!

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!