ChatGPT gets visual reasoning and full tool access

Together with

Welcome Humans🤖,

Here is what we have Today:

• 🧠 OpenAI's new o3 & o4-mini: The thinking machines.

• 💰 Google's AI thinking budget: Pay for smarter answers?

• 💻 1-bit wonders: Microsoft's AI runs on your laptop.

• 🧬 AI Just Rewrote the Rules of Protein Design.

• 💼 New Job Opportunities

OPENAI

🧠 OpenAI’s New Models Can Think in Images—and Code Like Pros

OpenAI just went full sci-fi with its latest drop — releasing o3 and o4-mini, its most advanced reasoning models yet. These aren't just faster or smarter — they think in pictures and can now fully operate ChatGPT’s entire toolbox.

And that’s not all. There's a new open-source coding agent too — and it lives in your terminal.

🚨 What’s Happening?

OpenAI just released two major updates:

• o3: Their smartest model yet, crushing benchmarks in coding, math, and science.

• o4-mini: A speedy, cost-efficient model outperforming older versions and acing exams like AIME 2025.

🛠️ Tools? They've Got 'Em All.

Both models now have full agentic access to every ChatGPT tool:

• 🌐 Web Search

• 🧮 Python

• 🎨 Image Generation

• 🤖 Custom Actions

They can combine these tools for next-level reasoning.

👁️ “Thinking with Images” is Now Real

These models don’t just read — they see.

They can now:

• Analyze visuals

• Manipulate images

• Integrate visuals into their decision-making

AI now thinks in pixels, not just text. 🔍🧠

💻 Meet Codex CLI

Also launched: Codex CLI

🧑‍💻 An open-source coding agent that runs right in your terminal.
🛠️ It bridges reasoning models with real-world coding tasks.

Think: ChatGPT, but hands-on in your dev environment.

🌍 Why It Matters

AGI isn’t just theory anymore.

“A GPT-4 level qualitative step into the future.”
— Greg Brockman, OpenAI President

With tool use + visual reasoning, these models are:

• Generating new scientific ideas

• Solving complex tasks with autonomy

• Climbing toward Step 4 on OpenAI’s AGI roadmap

We’re not just training chatbots anymore — we’re watching them evolve.

Together with The Rundown AI

Start learning AI in 2025

Everyone talks about AI, but no one has the time to learn it. So, we found the easiest way to learn AI in as little time as possible: The Rundown AI.

It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.

Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses – tailored to your needs.

Sign up to start learning.

AI Research

🧪 Biotech’s ChatGPT Moment Is Here

Profluent drops a breakthrough: ProGen3
A powerful new AI model that designs complex proteins from scratch — and reveals biology’s first scaling laws.

Profluent unveiled ProGen3, a powerful new family of AI models that can create entirely new proteins — and in the process, found the first evidence of scaling laws in biology.

Bigger models + more data = stronger protein design

🚨 What’s New?

🔬 46B parameter model trained on
📊 3.4 billion protein sequences
⚡️ Smashes previous records in dataset size & performance

🧪 What It Can Do:

🧫 Designed new antibodies
✅ Match real therapeutics in performance
🛡️ Different enough to dodge patent conflicts

🧬 Created mini gene editors
📉 Less than half the size of CRISPR-Cas9
🚀 Enables new gene therapy delivery methods

🎁 Bonus Drop: OpenAntibodies

💊 20 ready-to-license antibodies
🎯 Target diseases affecting 7 million+ people
💸 Available royalty-free or with upfront licensing

💡 Why It Matters

🧠 Turns biology into engineering
⚙️ Faster, more predictable, and scalable
🧨 Could revolutionize drug & gene therapy development
🌍 We’re just scratching the surface of AI’s impact on medicine

Trending Today

• Neo Gamma, A Human-like Humanoid developed by 1X technologies, designed humanoid Robots for home use featuring a soft knitted exterior and human like movement capable of performing household tasks and responding to voice commands walking with a natural human movement, like arm swinging when walking, squat down and even sit in a chair

• Google has launched Gemini 2.5 Flash, a lighter AI model that lets developers adjust how much reasoning the model performs via a token-based “thinking budget.” Reasoning can be capped at 24,576 tokens, with pricing ranging from $0.60 (reasoning off) to $3.50 per million tokens (reasoning on). The model balances speed and accuracy, performing well on GPQA and AIME benchmarks. It’s now in preview through Google AI Studio and Vertex AI.

• Microsoft has introduced BitNet b1.58 2B4T, a 2-billion parameter 1-bit AI model built to run efficiently on CPUs, including Apple’s M2. The model compresses weights into -1, 0, and 1, reducing memory usage and boosting speed. Trained on 4 trillion tokens, BitNet matches or exceeds similarly sized models like Meta’s Llama 3.2 1B and Google’s Gemma 3 1B on tasks such as GSM8K and PIQA. It runs up to twice as fast while using less memory.

• Meta’s FAIR research arm just published five new open-source AI research projects focused on perception and reasoning, showcasing advances in computer vision, 3D understanding, and collaborative AI capabilities.

Recommended Reading
If WE had to recommend other newsletters

AI Tool Report is an Newsletter, we enjoy reading daily. AI Tool Report delivers top techniques on how AI can transform your business filled with practical tips, real-world examples.

A Smart Bear is for people who love thinking about strategy, startups, product, marketing, decision-making, and founder psychology.

Job Opportunities

• Palantir Technologies - Deployment Strategist - Apply

• Shield AI - Manager, Frontend Engineering - Apply

• Findem - Search Optimization Associate - Apply

• DeepMind - Research Scientist - Apply

• DeepMind - Hardware Engineer, Design and Verification - Apply

AI News

Ideas? Comments? Complaints?

We read your emails, comments and poll replies daily.

Get the most important AI, tech, and science news in a free daily email.
New Here? Subscribe!
Sponsorship Slots Open for April and Reach over 800+ active readers. (Now 40% off) 🤯

Until next time, Stay Informed!