👀 Sam Altman FINALLY reveals GPT-5

Also: Google Gemini now remembers your Chats

Author

Evolving AI
February 14, 2025

Welcome, AI enthusiasts

Sam Altman just revealed OpenAI’s new model roadmap, confirming that GPT-5 will launch in a few months. Instead of a standalone o3 release, OpenAI is merging the O-series and GPT-series into one powerful system. Let’s dive in! 

In today’s insights:

  • OpenAI lays out plans for GPT-5

  • Google Gemini now remembers your chats

  • Adobe Firefly unveils first video generation model that it says is "safe to use"

Read time: 4 minutes

🗞️ LATEST DEVELOPMENTS

Source: Sam Altman on X

Evolving AI: OpenAI's CEO reveals future plans for GPT-4.5 and the awaited GPT-5.

Key Points:

  • OpenAI plans to make its product lineup simpler and it aims for a unified intelligence experience.

  • GPT-4.5, previously Orion, will be the last model of its kind before a big shift.

  • GPT-5 is planned to integrate various technologies offering different intelligence levels for users that are free Plus and Pro.

Details:

OpenAI CEO Sam Altman provided a glimpse into the future of the company's AI models models like GPT-4.5 and GPT-5. He addressed user confusion with the current product selection stating a commitment to simplification. The aim is a single more user-friendly intelligence experience. The upcoming GPT-4.5 will represent the final "non-chain-of-thought" model. GPT-5 signifies a move to bring together many existing capabilities including the "o3" system into one enhanced system.

Why It Matters:

We now have clarity on OpenAI's plan for GPT-5 and its arrival. GPT-4 was released in March 2023 so we are approaching two years without its successor. Sam Altman indicated GPT-4.5 is weeks away and GPT-5 months away so expect rising anticipation. Free ChatGPT users will experience GPT-5 features with unlimited basic access while paid subscribers will get enhanced features.

TOGETHER WITH VOICENOTES
🏆️ Voicenotes: Your 24/7 AI Second Brain

Evolving AI: This AI-powered voice-to-text tool allows you to effortlessly capture, organize, and retrieve your thoughts.

  • Available on mobile, watch, and web.

  • Capture fleeting thoughts & important conversations.

  • Effortlessly record and summarize meetings.

  • Instantly find anything you've ever said.

  • Join 150,000+ users.

Evolving AI: Google’s Gemini now recalls past conversations, making interactions smoother and more personalized.

Key Points:

  • Gemini AI can now summarize and recall details from past conversations.

  • The feature is available for English-speaking users on Google One AI Premium.

  • Users can manage or disable chat memory through privacy settings.

Details:

Google has introduced a memory feature for its Gemini AI chatbot, allowing it to recall details from previous conversations. This means users won’t have to repeat themselves or search through old threads for context. Initially available to English-speaking subscribers of Google One AI Premium, the feature will expand to more languages and enterprise accounts soon. Privacy controls let users review, delete, or turn off the recall function entirely.

Why It Matters:

This update makes AI interactions feel more natural and efficient. Users benefit from seamless conversations, reducing the need to re-explain details. Businesses using AI for customer service or workflow automation can expect improved responsiveness. However, stored chat history may still raise concerns for some.

Evolving AI: Adobe releases a new AI video tool trained on licensed data.

Key Points:

  • Adobe's Firefly Video model is now in public beta testing offering AI video tools trained using licensed content.

  • The system creates 1080P video clips from text inputs or image prompts and it works with Premiere Pro.

  • Tiered pricing plans provide different access levels starting at just $9.99 each month.

Details:

Adobe is bringing its generative AI Firefly to video. The new Firefly Video Model is commercially safe so it's trained with licensed content. This reduces intellectual property risk for all users. The model is available in a new web app and inside Premiere Pro too. The AI generates videos from text or images and it allows for control over things like camera angles and even special effects. Adobe indicates upcoming features like higher resolution and transparent video are coming. Pricing is structured to fit various user needs.

Why It Matters:

Adobe prioritizes a legally sound approach for its AI as they used licensed data for training so users won't face legal problems later. The Premiere Pro integration means that editors can refine AI content with the professional tools they already use. For instance Adobe showed adding snow to a scene and then instantly matching the color. This mix of AI help and professional control streamlines tasks as it lets creative people do more in less time.

 👀 Click on the image you think is real

🚀 Anthropic prepares new Claude hybrid LLMs with reasoning capability.

🔎 Reddit hints at expanded AI-powered search.

📱 YouTube is going to let you put AI-generated video clips in Shorts.

🆚 Thomson Reuters Wins Copyright Case Against Former AI Competitor.

🤨 UK drops ‘safety’ from its AI body.

👀 Baidu goes open source with Bernie AI.

📈 Trending AI Tools

  • 🚀 Taped - AI-powered workflow management (link)

  • 🎥 Spikes Crudio - AI-powered video editing tool (link)

  • 🧠 Mentalport - Streamline daily mental wellness (link)

  • 🔄 Lume - Automate data mappings with AI (link)

  • 📱Syllaby - AI-powered Social Media Strategy tool (link)

Reply

or to participate.