Howdy! It’s Barsee again.

Happy Wednesday, AI family, and welcome back to AI Valley.

In today’s edition:

• OpenAI just dropped a brand update

• ByteDance unveils new AI tool for highly realistic Deepfake videos

• AI creates a database to help detect cancer earlier

• Plus trending AI tools, posts, and resources

Ready, set, go…

OPENAI

OpenAI just dropped a brand update

OpenAI has introduced its first-ever rebrand, featuring a refreshed logo, updated color scheme, and a new custom typeface called OpenAI Sans, aiming to make its brand image feel more "organic" and "humanized".

Here's what you need to know:

• The original logo was designed by OpenAI's CEO Sam Altman and co-founder Ilya Sutskever, while the new design was led by an internal team headed by Veit Moeller and Shannon Jager.

• At the core of the rebrand is OpenAI Sans, a custom minimalistic typeface created for clarity and consistency.

• OpenAI’s designers revealed they took a hybrid approach, working with experts in photography, typography, motion, and spatial design while also using AI tools like DALL·E, ChatGPT, and Sora as creative partners.

• The revamped look has already been integrated across the ChatGPT interface, from web browsers to mobile apps, including the ChatGPT iOS app.

Why it matters:

While the changes are subtle, they reflect the company's move toward a more natural and human-centered image. OpenAI’s rebrand marks its evolution from a research lab to a product-focused company with offerings like ChatGPT and Sora. The new look gives the brand a warmer, more approachable feel while adding a playful touch that better aligns with its growing role in the tech ecosystem.

TOGETHER WITH WISPRFLOW

Work at the speed you think with Wispr Flow

Meet your new productivity hack: Work smarter, not harder with an app that transforms jumbled thoughts into clear communication. Flow captures your words in real-time and formats them so you don't have to. Flow automatically edits out filler words and clutter. Want a different tone or a quick AI-driven rewrite? Just ask Flow.

• 3x Faster Than Typing – Get more done in less time.

• Natural Voice – Flow sounds like you, not a stilted bot.

• Auto-Edits & Formatting – No more typos or punctuation errors.

• 100+ Languages – Communicate globally without missing a beat.

Because typing is so last decade. Think it, speak it, send it—Flow will handle the rest.

Try Wispr Flow and see how low effort writing can be.

SIDE UPDATES

🎭 ByteDance unveils new AI tool for highly realistic Deepfake videos

ByteDance demoed OmniHuman-1, an AI tool that generates full body videos of 5 to 25 seconds using a single image and audio clip. Trained on nearly 19,000 hours of human motion data, the tool produces lifelike videos of people speaking, singing, and performing other tasks. Unlike earlier models limited to facial or static full-body animations, OmniHuman-1 adapts to various aspect ratios and body proportions while accurately capturing natural facial expressions and detailed limb movements.

🤖 Figure AI leaves agreement with OpenAI, to show “major breakthrough [in] the next 30 days”

Figure AI's CEO Brett Adcock has promised to deliver something no one has ever seen in a humanoid robot in the next 30 days following a major breakthrough in the company's in-house AI. The news is a surprise, given the role that OpenAI plays in the cultural zeitgeist. In August, the two companies announced that the Figure 02 humanoid would use OpenAI models for natural language communication.

🗣️ OpenAI 1-800-CHATGPT on WhatsApp now supports Image and Voice notes

ChatGPT integration on WhatsApp is expanded with new capabilities. Users can now send audio messages instead of typing and receive text responses. You can also share images and ask follow-up questions for a more interactive experience. In addition, users can link their free, Plus, or Pro ChatGPT plans to maximize usage within WhatsApp. To get started, save 1-800-CHATGPT (1-800-242-8478) as a contact and relaunch the app.

🧬 AI creates a database to help detect cancer earlier

Johns Hopkins researchers developed AbdomenAtlas, an AI-powered dataset of 45,000 3D CT scans with 142 annotated structures, vastly improving medical image analysis for early cancer detection. It’s 36 times larger than its nearest competitor, compiled from 145 hospitals. AI and 12 radiologists completed in two years what would’ve taken 2,500 years manually, achieving a 500-fold speedup in organ annotation. Despite its scale, it represents just 0.05% of annual US CT scans.

🤗 Hugging Face develops ‘Open’ version of OpenAI’s Deep Research tool

Hugging Face has created an open-source version of OpenAI's Deep Research tool, completing the project within 24 hours of the original release. The new tool merges o1 AI model with an open framework, enabling it to navigate the web and read PDF files autonomously. While it performs well, it still trails behind OpenAI’s proprietary model in some benchmarks. The code is now available for public feedback on GitHub.

🕺 Meta introduces VideoJAM for realistic AI video motion

Meta has introduced VideoJAM, an AI model designed to eliminate the robotic motion often seen in AI-generated videos. Unlike traditional models that treat frames independently, VideoJAM focuses on both appearance and movement, resulting in smoother, more natural motion without requiring massive training data. In benchmarks centered on motion coherence, VideoJAM outperformed major models like Sora and Kling.

TRENDING TOOLS

• Natura > AI that remembers, adapts, and evolves (it introduces AI People, virtual assistants with memory, personality, and initiative.) *

• HuggingFace Spaces > AI App store that allows you to search approximately 400k AI tools for your best use.

• Tana - Put your notes to work with voice and AI.

• Replit Mobile App > Create apps and AI agents with text prompts directly from your phone.

• Teamsaver.ai - AI agent to collect daily team updates via email.

• Skyvern 2.0 > Build AI browser agents in plain english.

• Play AI Dialog 1.0 > An ultra-emotional AI text-to-speech model.

THINK PIECES / RESOURCES

• OpenAI’s Noam Brown called Deep Research the potential end of Wikipedia.

• Bill Gates on AI and Innovation.

• ChatGPT’s paid subs grew to about 15M in 2024. It was on 5.8M in 2023.

• The Vatican released guidelines for AI. It covers how to use across warfare, healthcare, and education, warning against replacing human judgment with machines.

• How I use LLMs as a staff engineer.

• AI prompt engineer’s best advice to get exactly what you want.

• OpenAI used this subreddit to test AI persuasion.

CONTENT CORNER

1/

2/

3/

4/

THAT’S ALL FOR TODAY

That’s all for today’s issue, folks.

💡 Help me get better and suggest new ideas at [email protected] or @heyBarsee

👍️ Like what you see? Subscribe here

Thanks for being here.

REACH 100K+ READERS

Acquire new customers and drive revenue by partnering with us

Sponsor AI Valley and reach over 100,000+ entrepreneurs, founders, software engineers, investors, etc.

If you’re interested in sponsoring us, email [email protected] with the subject “AI Valley Ads”.