This Week in AI: Essential Updates for Creators 4/13

Welcome to your weekly AI update, highlighting the most important news and tools for creators using AI.

1. ChatGPT Gets Memory – GPT-4o Now Remembers You

OpenAI rolled out memory for ChatGPT Plus users. This means GPT-4o can now remember your tone, your brand, and your style across chats. It’ll keep your preferences, tools, and context—making it more like a personal creative assistant.
You can also check and delete what it remembers anytime.

2. Google Drops Gemini 2.5 + Veo 2 for Creators

Google announced Gemini 2.5, its most advanced model yet—with better logic, context, and 1M token length. But what really matters: they launched Veo 2, a video generator that lets you create 720p AI videos from text prompts. You can also resize, change backgrounds, and generate visual assets.

Perfect for creators wanting cinematic shots, fast reels, or vertical video.

3. Midjourney V7 Launches – Way Better Visuals

Midjourney V7 is now in alpha and producing way more accurate hands, faces, and camera detail than previous versions.
It also has personalized prompt tuning—so the more you use it, the more it adapts to your style.

Draft Mode = faster/cheaper test images too.
This is now one of the strongest tools for thumbnails, story slides, and branded visuals.

4. Adobe Adds Generative Video Tools to Premiere Pro

Adobe launched “Generative Extend” in Premiere Pro, letting you extend footage and ambient audio using AI—perfect for when your video’s 1 second too short or you need more room for a text overlay.
Also, new AI tools in Photoshop (Firefly 3 beta) improve text generation, object removal, and background fills.

5. YouTube Is Rolling Out AI-Powered Caption Translations

YouTube Shorts and long-form videos are starting to support auto-translated captions, powered by AI. This means your content could soon reach global viewers automatically—no need for separate translated uploads.

If you're growing internationally or posting multilingual content, this is huge.

6. GPT-4o Image Generation Prompts Now Deliver More Detail

Prompt structure matters more than ever. GPT-4o’s image generation has gotten more accurate with:

• Scene control (lighting, depth, tone)

• Layout consistency

• Text placement

• Prompt length (longer = better)

If you’re making visuals inside ChatGPT, this is the time to use structured prompt packs and refine your inputs.

That’s everything you should know this week — the stuff that actually matters if you’re using AI to create.
See you next Saturday with more.

Best,
Luke