• AI Valley
  • Posts
  • Google enters the humanoid game

Google enters the humanoid game

OpenAI to release o3, Google unveils Gemini 2.0 Flash Thinking...

Together with

Howdy! It’s Barsee again.

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

In today’s edition:

  • Google DeepMind enters the humanoid game

  • OpenAI to release ‘o3’ soon

  • Plus trending AI tools, posts, and resources

Ready, set, go…

GOOGLE ROBOTICS

Google DeepMind enters the humanoid game

(Image of Apptronik Robot)

Google DeepMind and Apptronik have joined forces to develop humanoid robots designed for dynamic, real-world environments. The partnership leverages DeepMind's expertise in AI and machine learning alongside Apptronik's decade-long experience in human-centered robotics design.

Here's what you need to know:

  • Apptronik, founded in 2016, has developed 15 robot models, including NASA’s Valkyrie, and is now focusing on its latest Apollo robot, standing 5’8 tall and weighing 73kgs.

  • Apollo is already used in Mercedes-Benz manufacturing facilities for repetitive tasks like delivering parts and inspecting components.

  • DeepMind’s Gemini AI models will enhance Apollo’s capabilities, integrating reasoning and real-world action for general-purpose tasks.

  • A recent demo showed Apollo using Gemini’s spatial reasoning to delicately play Jenga, guided by a simple text prompt.

Why it matters:

The partnership between Google DeepMind and Apptronik is only the second of its kind, following Figure's collaboration with OpenAI, highlighting a growing trend of pairing advanced AI with cutting-edge robotics.

TOGETHER WITH ALPHA

This AI tool gives you real-time investment insights

(Image credit: Alpha)

Meet Alpha: Your AI-powered investment watchlist, now available as a standalone app for iOS devices. Originally built by the investing platform Public, Alpha actively monitors the stocks, ETFs, and crypto you care about, delivering real-time alerts on breaking news, market movements, earnings calls, and more.

When an asset on your watchlist changes, Alpha gives you the why behind the move with clear, contextual insights that make sense of market activity. Plus, you can ask Alpha any question using its intuitive natural-language interface to dive into market trends, earnings, or historical data.

Download Alpha now on iOS, and start building your investment watchlist for just $1/week—or free if you're a Public member.

SIDE UPDATES

(Gif credit: OpenAI)

ChatGPT now connects with popular apps like Apple Notes, Notion, Xcode, and VS Code, making it easier to automate tasks through voice or text. The new "Work with Apps" feature, available only on macOS for now, also includes an Advanced Voice mode for seamless interaction. Windows users and free accounts will get access in 2025.

Google’s Gemini 2.0 Flash Thinking model is designed for advanced reasoning across coding, math, and physics. It uses text and visuals to solve tough problems and is ideal for multimodal tasks. Early testers can access it now via Google AI Studio or the Gemini API as part of its experimental rollout.

Researchers at Memorial Sloan Kettering have developed AI-powered nanotube blood tests that detect diseases like ovarian cancer by analyzing molecular markers. The nanotubes emit light that AI interprets for early warning signs. Initial tests show it’s more accurate than traditional methods, with great potential for saving lives.

Meta is bringing AI video editing to Instagram next year, powered by its Movie Gen model. Users can tweak outfits, backgrounds, and more with simple text commands. Instagram head Adam Mosseri showed off the tool by turning himself into a puppet, showcasing its creative possibilities for users.

Perplexity AI has raised $500 million in a funding round, tripling its valuation to $9 billion in just six months. The company focuses on AI-enhanced search tools, which are gaining traction as alternatives to traditional search engines. Investors see massive growth potential in this space.

OPENAI O3

OpenAI to release ‘o3’ soon

OpenAI is reportedly developing its next-generation "o3" reasoning model, which will take more time to “think” about user queries before responding.

Here's what you need to know:

  • OpenAI opted to skip the name “o2” to avoid potential trademark conflicts with O2, a British telecommunications service provider.

  • The "o3" model is designed to improve upon o1’s reasoning capabilities, aiming for more natural and logically sound responses during interactions.

  • After facing a slowdown in gains from using more compute and data during pretraining, OpenAI invested heavily in its reasoning AI research and used a new pretrained model, Orion, to develop the o3 model.

Recently, Sam Altman hinted at the release of the next model on X with a playful “oh oh oh,” likely referring to o3.

(Source: Sam’s post)

Let's solve this:

  • ho ho ho = (h + o)(h + o)(h + o)

  • Since "h" is silent in Christmas:

  • o o o = o³ = o3

Keep an eye out for the official launch, it looks like it’s going to launch in coming days.

TRENDING TOOLS

  • Gemini Code Assist > Instantly completes code and generates full functions on demand. (link)

  • Trickle > Build stunning AI Apps, Websites, and Forms with ease. (link)

  • Tempo > Visual editor for React, built for developers and designers. (link)

  • GenFuse AI > Automate any work with AI agents. No technical skills needed. (link)

THINK PIECES / RESOURCES

  • AI's impact on industries in 2025. (link)

  • How to deploy your own Gemini Multimodal Live voice agent that you can call on the phone (or that can call you) in 5 minutes. (link)

  • The promise and warning of Truth Terminal, the AI bot that secured $50,000 in bitcoin from Marc Andreessen. (link)

  • How to create GPT to take meeting notes. (link)

  • Building effective agents - Anthropic Research. (link)

  • Best-of-N (BoN) Jailbreaking, a simple black-box algorithm that jailbreaks frontier AI systems across modalities. (link)

  • Learn how to effectively prompt and use OpenAI’s o1 model. (link)

CONTENT CORNER

Waymo's self-driving cars are advancing at an incredible pace, and in 2025, Tokyo will become the first city outside the U.S. to test these autonomous taxis.

Every level of Internet explained in 7 minutes.

How bots use YT comments to find private info.

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”.