Beginners in AI

Thank you for joining us again!

Welcome to this week's edition of Beginners in AI, where we explore the latest trends, tools, and news in the world of AI and the tech that surrounds it. Like all editions, this is human curated and published with the intention of making AI news and technology more accessible to everyone. 

This week, OpenAI launches powerful new O-series models, introduces a cost-saving Flex tier, and rolls out biosafety safeguards. Google gives college students free access to Gemini Advanced, Anthropic enhances Claude with new tools, and VoicePatrol debuts live voice protection for gamers. Plus, Harvard’s RoboBee gets nature-inspired legs for smoother landings.

AI TOP STORY
OpenAI Unveils O3 and O4 Mini AI Models

A New Generation of AI Smarts
OpenAI has quietly launched two new models, o3 and o4‑mini, able to read images and call digital tools in a single flow. Internal testers say both variants already live in ChatGPT Plus and select enterprise pilots, giving the company real‑world feedback before a broader release. Early benchmarks point to strong gains on multimodal reasoning challenges such as MMMU and MathVista.

Smarter, Leaner, More Capable
The duo sits inside OpenAI’s move toward agentic AI, where a single prompt can trigger multi‑step actions over APIs or desktop apps. According to the official System Card, o4‑mini scores 99.5 % on the 2025 AIME math contest when it can run Python, and it runs on fewer flops than GPT‑4. The smaller footprint lets the model serve customers at lower latency and cost, yet still outperforms GPT‑4 on many perception tasks. Engineers who have tried it note that hallucination rates remain higher than earlier GPT‑4o builds, so tooling for fact‑checking is part of the recommended stack.

Why It Matters to You
Agent‑style models can take chores off your plate. Instead of giving answers and leaving you to copy the result, ChatGPT with o3 or o4‑mini could open a spreadsheet, fill cells, rename files, or queue image edits, then report back. For busy teams, that means project hand‑offs to software rather than coworkers and a shorter time from idea to output — a step toward truly hands‑free smart assistants.

Read More

LAST WEEK IN AI AND TECH

Flex Appeal: OpenAI Unveils Budget-Friendly AI Mode
OpenAI has launched “Flex,” a new compute tier for users who don’t need lightning-fast responses from their AI tools. By scheduling slower inference times, users can save on cost, making large-scale usage (like batch document processing) much more affordable. It’s a move that could appeal to businesses seeking scalable AI without breaking the bank. Flex won’t always use the latest model, but it provides consistent quality for non-urgent tasks — a big win for price-conscious users.

Read More

Gemini Grants: Google One Gives AI Tools to Students
Google is offering its AI-powered Gemini Advanced features for free to college students, normally part of the $20/month AI Premium plan. Students with .edu emails can now access tools like document summarization, image generation, and tutoring help at no cost. This strategic move expands Google’s footprint in the education sector while giving future professionals hands-on experience with Gemini’s capabilities. It's also seen as a play to outpace Microsoft and OpenAI in classroom adoption.

Read More

Bio Blocker: OpenAI Models Gain Safeguards Against Dangerous Science
In response to increasing concerns around AI being used to generate dangerous biological weapons, OpenAI has embedded a new safety system into its models. This system restricts access to potentially risky biochemical or bioengineering prompts, especially those related to pathogens or toxins. The change came after OpenAI collaborated with experts in biosafety and dual-use research. It’s a sign that AI labs are proactively addressing real-world misuse scenarios as their models grow more capable.

Read More

Mic Checkmate: VoicePatrol Guards Game Chats in Real Time
VoicePatrol has introduced an AI tool that protects gamers from real-time voice harassment. The system uses speech recognition and sentiment analysis to detect toxic behavior as it happens, rather than flagging it after the fact. Designed for integration with multiplayer games, it offers in-game moderation that’s both fast and adaptive. This could transform online gaming into a safer, more inclusive space — especially for younger players and streamers. Notably, the system is supposed to still allow users to engage in non-PG banter and avoid micromanagement.

Read More

Claude in the Cloud: Anthropic Adds Research Tool & Google Workspace Integration
Anthropic is expanding Claude’s capabilities with a new research interface and deep integration into Google Workspace apps like Docs and Sheets. The research tool is aimed at enterprise users needing fast, structured synthesis from long or complex documents. Meanwhile, the Google tie-in means Claude can now assist directly inside productivity tools — helping users write, edit, and analyze without leaving their workflow. It’s another signal that AI is becoming a background layer across everyday apps.

Read More

TECH TERMS TO KNOW

Federated Learning: A method for training AI models across many devices (like smartphones), so each device’s data stays local and privacy is preserved while still improving a global model

TOOL SPOTLIGHT (non-sponsored)

Clueso is an AI-powered software platform designed to transform simple screen recordings into professional-quality videos, step-by-step guides, and comprehensive documentation. It automates the creation of training materials, product demos, onboarding guides, and knowledge base articles, making it especially valuable for teams that need to produce clear, branded, and multilingual content quickly and at scale.

Benefits:

• Massive Time Savings: Clueso automates content creation, reducing production time from hours to minutes.

• AI-Enhanced Professional Quality: It uses AI to generate polished scripts, studio-quality voiceovers, and branded videos in multiple languages.

• Seamless Updates and Collaboration: Easily update documentation and collaborate with your team without starting from scratch.

• All-in-One Platform with Enterprise Features: Clueso offers a secure, searchable knowledge base with analytics and integration options.

• User-Friendly and Highly Rated: The platform is intuitive, customizable, and highly rated by users for its efficiency and support.

ROBOTICS AND AI

Tiny Touchdown: RoboBee Learns Graceful Landings from Nature

Inspired by the crane fly, Harvard researchers have outfitted the miniature flying robot “RoboBee” with flexible, springy legs that enable it to land softly on a variety of surfaces. Previously, the microrobot's high-speed wings made landings jarring and unstable. With the new limbs, RoboBee can now perch gently — even on uneven terrain — increasing its stability and reusability in real-world environments.

This biomimetic upgrade means RoboBee could one day be deployed in delicate environments like disaster zones, forests, or even indoor spaces, offering a lightweight, minimally invasive solution for surveillance, research, or search-and-rescue tasks.

Read More

TRY THIS PROMPT (copy and paste into ChatGPT or Grok)

Personal movie poster for yourself, friends, and family. 

After you upload your photo, just paste this prompt (no other edits needed except swapping in your movie title):


Generate a movie‑poster‑style illustration inspired by “<Your Favorite Movie Title>.”  
– Use the uploaded image (referenced_image_ids:["<your_image_id>"]) as the main character.  
– Match the film’s signature color palette and lighting mood.  
– Render the title at the top in the movie’s original font style.  
– Add a fitting tagline beneath the title in that same font.  
– Surround the figure with key visual elements from the official poster (for example, lens flares, silhouettes, smoke).  
– Place a block of mock credits and a faux release date at the bottom in the poster’s usual layout.


How to use  
1. Start a new ChatGPT image request.  
2. Upload your photo.  
3. Copy‑paste the prompt above.  
4. Replace `<Your Favorite Movie Title>` with your chosen film, and `<your_image_id>` with the ID shown after your upload.  
5. Hit “Generate” and voilà—your photo transformed into a cinematic poster!

DID YOU KNOW?

In October 2019, Google’s 53‑qubit Sycamore quantum processor completed a sampling task in just 200 seconds—a computation that the world’s then‑fastest classical supercomputer would have taken an estimated 10,000 years to finish.

30 Referrals: Lifetime access to all Beginners in AI videos and courses

AI-ASSISTED IMAGE OF THE WEEK

Interested in stock trading, but no clue where to start? Sign-up for our sister newsletter launching daily in April. Each day will build onto the next using current news as a learning aid.

Thank you for reading. We’re all beginners in something. With that in mind, your questions and feedback are always welcome and I read every single email!

-James

By the way, this is the link if you liked the content and want to share with a friend.