ChatGPT can now remember everything

Hello pals! Here’s what you need to know about AI today:

👉 OpenAI adds enhanced memory to ChatGPT

👉 Canva launches an AI assistant to add AI to its graphic design platform

👉 Google wants to merge Veo AI video generator and Gemini

and many more!

📧 Did someone forward you this email? Subscribe here for free to get the latest AI news everyday!

Read time: 5 minutes

What’s going on: OpenAI has upgraded ChatGPT’s memory feature to reference past conversations, enabling more tailored and context-aware responses. Unlike the previous system, which relied on users explicitly prompting the AI to save specific details, the new functionality automatically draws from all prior chats to adapt to a user’s tone, preferences, and interests.

What does it mean: This new feature aims to make interactions feel more natural, as if ChatGPT is building on an ongoing relationship rather than starting fresh each time. The upgrade applies across text, voice, and image-generation features, making the AI more versatile in tasks like writing, learning, or seeking advice.

More details:

• The rollout prioritizes ChatGPT Pro and Plus subscribers, with Enterprise, Team, and Edu users gaining access in the coming weeks.

• Users in the UK, EU, Iceland, Liechtenstein, Norway, and Switzerland are excluded for now due to regional data privacy regulations requiring further review.

• OpenAI emphasizes user control, allowing people to opt out of memory referencing entirely or disable it for specific chats via a Temporary Chat mode, which functions like an incognito session.

• Users can also ask ChatGPT what it remembers or delete specific memories, though the chat history itself remains stored unless manually cleared.

• For a detailed deep dive, read our article on The 79.

What’s going on: Canva, the famous online graphic design platform, has launched Visual Suite 2.0, a major update that integrates AI-driven tools, coding capabilities, and spreadsheets into its design platform. The centerpiece is Canva AI, a conversational assistant powered by Anthropic’s technology, which handles tasks like generating images, writing copy, and creating documents based on user prompts.

What does it mean: This new AI assistant aims to streamline workflows by offering design suggestions and producing content tailored to specific needs, such as social media graphics or print mock-ups. The platform’s new features target both individual creators and businesses, expanding Canva’s scope beyond its traditional graphic design roots.

More details:

• Canva Code lets users create interactive elements like maps or calculators without coding expertise. By entering a text prompt, anyone can generate mini-apps to embed in designs, making websites and presentations more dynamic.

• These features draw from Canva’s internal prototyping experiments, now extended to its 230 million monthly users.

• Canva Sheets introduces a visual take on spreadsheets, integrating data from tools like Google Analytics and HubSpot. AI-powered features like Magic Insights analyze data for patterns, and Magic Charts transform numbers into branded visuals, simplifying the creation of reports and infographics.

• For businesses, Visual Suite 2.0 builds on Canva’s enterprise offerings, adding tools like single sign-on and data integrations to support larger teams.

• Want to try out this brilliant product? Check it out on Canva’s website.

📽 Google DeepMind CEO Demis Hassabis announced plans to merge the Gemini AI models with Veo video-generating models to enhance Gemini's understanding of the physical world.

💰 Mira Murati's new AI startup, Thinking Machines Lab, is reportedly seeking a $2 billion seed round, which would value the company at least $10 billion, attracting investors with its team of high-profile AI researchers despite having no current product or revenue. Mira Murati is the ex-OpenAI CTO.

📅 OpenAI will retire GPT-4 from ChatGPT on April 30, replacing it with the superior GPT-4o, while GPT-4 will still be available via OpenAI's API.

🎵 YouTube is rolling out a "Music Assistant" tab in Creator Music that lets creators generate instrumental backing music using AI prompts, offering a free, copyright-safe option for adding music to videos; this feature is initially available to US creators in the YouTube Partner Program.

🤖 The next major ChatGPT release, potentially named GPT-4.1, could launch as soon as next week, featuring smaller "mini" and "nano" versions and major memory upgrades.

⚡ Google and PJM are partnering with Alphabet's Tapestry to develop AI models that will streamline the application process for connecting new power sources to the grid, which could help alleviate concerns about underpowered data centers and promote the integration of renewables, as 2.6 terawatts of generating capacity are currently awaiting approval nationwide.

🚀 Deep Cogito, a startup founded by ex-Google AI employees, has launched a family of hybrid AI models called Cogito 1 that can toggle between reasoning and non-reasoning modes, outperforming existing open models of similar size in internal benchmarks.

AI + Automation with Python

Pattern Data - Senior Data Engineer

HubSpot - AI, Senior Product Designer II

Contentstack - AI Intern

Logic20/20 - Senior Machine Learning Engineer

Thank you for staying with us like always! If you are not subscribed, subscribe here for free to get more of these emails in your inbox! Cheers!