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.

Our top story this week is all about ChatGPT’s new agent mode that can plan, shop and perform actions within its own browser for you. Microsoft’s 3‑D experiment, Figma’s Make tool, Brave’s move to block Windows Recall, a (not)surprising turn in chip politics, new personality settings for ChatGPT, and a warning about privacy in chat logs. Our robotics spotlight Amazon’s millionth robot and a traffic‑smart system for warehouse machines.

AI TOP STORY
Open-AI Releases Agent Mode

What Happened

OpenAI introduced a general‑purpose agent mode in ChatGPT on July 17. The company says the agent can navigate a calendar, make slide decks and run code, turning ChatGPT from a chat partner into a helper that finishes tasks. The tool merges features from earlier projects, blending Operator’s remote browser with Deep Research’s web synthesis. People who subscribe to the Pro, Plus or Team plans can turn on “agent mode” in the app. In a post announcing the launch, OpenAI wrote that “ChatGPT can now do work for you using its own computer … introducing ChatGPT agent — a unified agentic system combining Operator’s action‑taking remote browser, deep research’s web synthesis, and ChatGPT’s conversational strengths.” The agent can link to services like Gmail or GitHub and can carry out complex projects such as buying ingredients for Japanese breakfast or analyzing competitors.

What It Means

This new mode signals a shift in how people might use ChatGPT: instead of typing a question and reading an answer, they can delegate routine digital chores. Early in 2025, OpenAI debuted Operator, a research preview that booked travel and shopped online. The new agent brings those abilities together in one place and can tackle harder jobs like planning a meal and making a presentation. The quote above underscores OpenAI’s plan to give ChatGPT a “unified agentic system.” At the same time, the feature is restricted to paying users and still requires confirmation before it performs actions like buying or deleting data. OpenAI’s upcoming GPT‑5 model is expected to include mini and nano versions; pairing those lighter models with the agent could one day let it work even when your device has no web connection.

What to Take Away

Rapid progress in language models makes it tempting to assume a straight path to artificial general intelligence, yet each release still has gaps. Altman’s caution about server capacity is a reminder that computing power and energy remain real limits. The prospect of running smaller versions on personal devices could widen access but also raises questions about responsible use. Readers can watch for how GPT‑5 handles nuance and whether it narrows the gap between AI and human judgement.

Read More

LAST WEEK IN AI AND TECH

3‑D Snap! – Copilot turns pictures into models

Microsoft is experimenting with a Copilot tool that turns photos into three‑dimensional objects. In the test, users can upload a picture, select a region and press a “Recreate” button; the new object then appears under a “My Creations” tab, and “it is possible to generate a 3D object from an image, as well as to browse the ‘My Creations’ tab”. Reviewers note that the output is still rough and that there’s no clear timeline for public release. Even so, the feature hints at a near‑future where simple snapshots become 3‑D assets for games or design. For now the project remains an experiment as Microsoft gathers feedback.

Read More

App Magic – Figma’s Make tool opens to everyone

Figma announced that its AI‑powered Make feature is now available to all users, not just full‑seat subscribers. The tool turns natural‑language prompts into working app prototypes and lets designers upload reference images; Figma says the bundled AI credits will “easily cover day‑to‑day needs” and that teams can buy more later Only full seats can publish prototypes, which may encourage upgrades, yet everyone can now experiment with building an app by simply describing it. This shift lowers the barrier for people who want to sketch an idea without learning complex design software. Early adopters say Make helps them turn ideas into screens in minutes.

Read More

Privacy Shield – Brave disables Windows Recall

Browser maker Brave will disable Microsoft’s Recall feature from capturing snapshots inside its browser by default. Recall is a Windows function that takes screenshots every few seconds and stores them locally for AI‑powered search. Explaining the move, Brave’s vice‑president of privacy wrote that Recall “is antithetical to Brave’s goals as a privacy‑first browser,” warning that browsing activity should not be placed in a persistent database. The block underscores ongoing concern about recording users’ screens without clear consent. For users, it means Brave will act as a protective layer against software that remembers everything on screen.

Read More

Chip Drama – Trump rethinks breaking up Nvidia

Former U.S. president Donald Trump recently admitted that he once wanted to split Nvidia but changed his mind after learning how dominant the company is in AI chips. In his words, an adviser told him that Nvidia “has essentially 100 percent of the AI chip market,” so breaking it up could create a mess. He also praised CEO Jensen Huang and suggested he sees more value in the firm working closely with American counterparts. The comments imply that antitrust action against Nvidia is now less likely, even though regulators have been examining the company’s pricing. Observers note that political views on chip makers can shift quickly when leaders look at supply‑chain realities.

Read More

New Attitudes – ChatGPT gains personalities and Study mode

OpenAI quietly added personality options to ChatGPT and a “Study Together” mode. In the customise menu, users can choose settings such as Robot (direct and literal), Cynic (sarcastic), Listener (empathetic) or Sage (wise) — the “Robot” setting responds “in a direct, no‑nonsense manner”. The Study Together mode guides learners through topics, offering explanations, practice problems and quizzes. These additions hint at an effort to make chatbots more personal and educational. For example, a student can ask for help with algebra and receive step‑by‑step guidance instead of a simple answer.

Read More

TECH TERMS TO KNOW

Recall (Windows) – A Windows feature that automatically captures frequent screenshots of a user’s activity so an AI assistant can search past events.  Example: Searching for an image of a recipe you looked at yesterday by asking the assistant, which then scans stored screenshots to find it.

TOOL SPOTLIGHT (non-sponsored)

Podscastle is an all-in-one platform designed for creators producing long-form content, such as podcasts and videos. It offers a suite of tools to streamline production, editing, and distribution, focusing on accessibility, ease of use, and professional-quality content.

Why Podcastle Might Be Worth Trying

• Effortless recording & editing: Users can record audio and video easily, then edit quickly with modern, AI-powered tools.

• AI tools for efficiency & quality:

  • One-click background noise removal and audio enhancement.

  • AI-powered background blurring for a polished video appearance.

  • Automatic removal of silences, "umms," "ahhs," and other filler words.

  • AI-generated captions to improve accessibility and engagement.

  • Maintain simulated eye contact in recorded video.

• Text-based content creation: Turn typed text directly into lifelike audio, and seamlessly create audio from transcripts.

• Collaboration: Teams can work together efficiently in a shared, secure workspace.

• Content repurposing: Convert audio to text for accessibility or multi-platform publishing.

• Strong user endorsements: Numerous testimonials highlight time savings, user-friendliness, and professional results.

ROBOTICS AND AI

Amazon announced that it has deployed its one‑millionth mobile robot across its fulfillment network. Vice‑president of robotics Stefano La Russa said, “We’ve just deployed our 1 millionth robot,” highlighting the company’s scale. Amazon also revealed DeepFleet, a generative AI system that acts like a traffic director for its robots. The model coordinates robotic movements in warehouses, reducing travel times by roughly 10 percent and easing congestion. DeepFleet learns the layout of each facility and decides when robots should slow, speed up or pause to avoid bottlenecks. The milestone demonstrates how AI manages physical operations at a scale that few companies can match.

Read More

TRY THIS PROMPT (copy and paste into ChatGPT, Grok, Perplexity, Gemini)

Futurist Economist Edition

"Map the economy of a world run on code and creativity."

“Act as a futurist economist in the year 2045. Describe the global economy now that AI handles most production and routine services. Provide: (1) three core industries that have emerged to replace traditional jobs (e.g., memory curation, digital habitat design, emotional experience markets); (2) how governments tax and regulate AI-driven productivity; (3) what the ‘currency’ of this era looks like—tokens of time, attention, or something stranger; (4) two major challenges society faces in wealth distribution and inequality; and (5) an illustrative ‘day in the life’ vignette of an average citizen earning, spending, and bartering in this future system.”

DID YOU KNOW?

Amazon’s was using robots since at least 2012. When Amazon bought Kiva Systems in 2012, it began using squat orange robots to bring shelves to human workers.

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.