Hello and welcome to our weekly roundup!
Hope your week was a good one. Let's close with some great news and events.
Today, we have a small but extremely important update to ChatGPT, impressive progress on video generation, a few huge (but potential) deals, and more. And, of course, additional tools, guides, and bonuses!
Coincidentally, early access to Manus finally reached us this week. Check the main features of this agent and why it deserves attention here:
This Creators’ AI Edition:
Featured Materials 🎟️
News of the week 🌍
Useful tools ⚒️
Weekly Guides 📕
AI Meme of the Week 🤡
AI Tweet of the Week 🐦
(Bonus) Materials 🎁
Your advertisement could be featured here!
Sponsor a spot in our newsletter to connect with subscribers interested in AI, technology, and startups. For inquiries, click the button below.
Featured Material 🎟️
ChatGPT Gets Memory
OpenAI has rolled out a long‑term memory upgrade for ChatGPT.
The update lets the chatbot reference all past conversations; even those users didn’t save, to deliver responses that align with their preferences. This feature promises to make communicating with AI smoother and more tailored.
With this update, ChatGPT now draws on both user‑saved memories and an internal history of every chat session. OpenAI said the upgrade aligns with its vision of AI systems that know users over their lifetimes and can act as personal assistants.
The feature is already available for Plus and Pro subscribers. That said, the update is unavailable in regions with strict AI rules, including the European Union, the UK, Switzerland, Norway, Iceland, and Liechtenstein.
The new memory system works in two modes. In the first, ChatGPT stores details that users explicitly ask it to remember, such as personal preferences or project notes. In the second, the bot draws insights from chat history to fill in context without manual prompts. Users remain in control.
You can toggle off saved memories or chat history in the Personalization settings. For conversations that must remain ephemeral, a Temporary Chat option prevents any memory retention. If memory is off, ChatGPT will not access past chats by default.
Keep your mailbox updated with key knowledge & news from the AI industry
OpenAI is not a pioneer in this endeavor. The update follows similar moves by rival AI firms. In February, Google added a recall feature to its Gemini AI, and Anthropic said it will add memory to its Claude chatbot later this year. Thus, companies want to bring their models closer to the status of really useful and aware assistants.
While this is a minor update (by the standards of a startup that releases new models almost every month), it's essential for all of us.
We are witnessing a shift from episodic interactions with AI to a state of permanent dialog. This means less prompt engineering and tedious formulations but more results and benefits. Very cool feature.
News Of The Week 🌍
Nvidia & Stanford Demo 1‑minute AI Video Generation
Nvidia and Stanford have developed Test-Time Training, a new model that can generate coherent, minute-long videos from a single prompt. Unlike previous systems that collapse after a few seconds, this one holds character identity and motion across full scenes. Developers have also trained their AI on the Tom & Jerry series and made new episodes.
Test-Time Training is available on GitHub. And you can learn more about the model and prompts for generation one-minute long videos here.
OpenAI Discussed Buying Jony Ive’s AI Startup
According to The Information, OpenAI was considering acquiring Johny Ive’s (former Apple design guru) hardware startup. Details of the potential deal were not disclosed, but insiders say the estimated deal was worth $500M. What exactly Ive's startup is developing is also a mystery. Although it is known that it is about a physical device with AI.
Meta Launches Llama 4 AI Models
Meta has unveiled its latest open-source AI models under the Llama 4 series, featuring two initial releases—Llama 4 Scout and Llama 4 Maverick—and a third, Llama 4 Behemoth, still in training. These models introduce a Mixture of Expert (MoE) architecture for improved efficiency and performance.
Scout offers a 10M token context window, while Maverick, with 400B parameters, excels in multilingual and general assistant tasks. The models are accessible via Hugging Face and Meta's apps.
Sharing is caring! Refer someone who recently started a learning journey in AI. Make them more productive and earn rewards!
Canva Rolls Out AI‑Driven Visual Suite 2.0
At Canva Create 2025, the company unveiled Visual Suite 2.0, an AI-based unified design platform. It embedes AI across spreadsheets, content generation, data visualization, code creation and photo editing to accelerate creative workflows. The update also includes an AI companion that accepts voice, text or media prompts to brainstorm and refine designs in real time, and a revamped Photo Editor with an AI Background Generator.
Anthropic Launches $200/mo Claude Max Plan
Anthropic looked at the success of OpenAI's ChatGPT Pro and created a similar subscription. The startup has launched a Max plan for the most demanding users who use Claude on a regular basis. Subscribers will get access to all released models and features, and will be first in line to test the next updates.
Meanwhile, unlike ChatGPT Pro, Max offers two tiers: for $100/mo and $200/mo.
Mira Murati’s Startup Reportedly Raising $2B
Business Insider reporters have revealed that startup Mira Murati, OpenAI's former chief technology officer, is in talks for a really big deal. According to the leak, Talking Machines could raise $2B at a valuation of at least $10B. With the money, the startup promises to make AI systems “more widely understood, customizable and generally capable.”
This deal will be the largest seed round in history if the leak is true. Even by the standards of AI, which is attractive to investors, this is impressive money.
Useful Tools ⚒️
Writehere.site - An open framework for human-like long-form writing
self.so - Create a personal site in 30 seconds from your resume or LinkedIn
GitSummarize - Create quick-to-read docs from any GitHub repo
Crono – Outreach personalization at scale with data and AI
Command Bar – Turn your user's words into actions
Command Bar is an interface that lets you describe actions in plain language, which the system then executes within your product. It supports multi-step workflows, integrates with any web framework via a type-safe SDK, and keeps your product logic and data isolated from the AI layer.
It's designed to be integrated in under 15 minutes and is well-suited for SaaS products aiming to enhance user experience with natural language interfaces.
Share this post with friends, especially those interested in AI stories!
Weekly Guides 📕
AI Vibe Coding Tutorial + Workflow
25 ways to use AI in your app (and no hype)
This changes work forever – Write & research 10x Faster
I Built an AI Agent That Generates 50,000 Leads a Month (Here's How)
AI Meme Of The Week 🤡
AI Tweet Of The Week
A great example of a great use case.
(Bonus) Materials 🏆
How Google Used AI to Re-Create ‘The Wizard of Oz’ for the Las Vegas Sphere
The AI Agent Era Requires a New Kind of Game Theory
Professor Answers AI Questions | Tech Support | WIRED
Is Human Data Enough? With David Silver
Anthropic Education Report: How University Students Use Claude
OpenAI’s CPO on how AI changes must-have skills, moats, coding, startup playbooks, more
Share this edition with your friends!