Google is cooking AI

PLUS: Meta stops hiring

Author

Barsee
21 Aug

Together with

Howdy. It’s Barsee again.

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

Today’s climb through the Valley reveals:

  • Google Gemini to get text-to-image generations powered by nano-banana

  • OpenAI crosses $12B ARR in July, but faces a compute bottleneck

  • Google announces its AI-powered Pixel 10 smartphones

  • Harvard dropouts to launch ‘always on’ AI smart glasses

  • Plus trending AI tools, posts, and resources

Let’s dive into the Valley of AI…

SUPERHUMAN AI

Image: Superhuman

Most inboxes slow you down.

They bury what matters, surface what's trivial, and keep you reacting all day.

Superhuman is the opposite with AI that acts like your best assistant. It auto-labels, archives noise, reminds you before you forget, and drafts replies in your voice, instantly.

Imagine getting control, clarity, and speed every time you open your inbox.

*This is sponsored

THROUGH THE VALLEY

Google is preparing to roll out its nano-banana AI model across creative tools, with early signs pointing to integration in Gemini, Whisk, and Flow. In Flow, the model will power a new text-to-image feature where users can upload reference images and generate multiple variations for video projects. Google is also testing vertical aspect ratios for TikTok and YouTube Shorts, plus “preamble” presets like Vlogging or Cinematic that expand prompts into richer text. Social features such as video favoriting and QR sharing are coming too. While no launch date is set, these updates highlight Google’s push to unify image, video, and creative workflows in one platform.

OpenAI CFO Sarah Friar says the company’s biggest challenge isn’t revenue but the huge demand for computing power. Despite hitting milestones like $10 billion in annual recurring revenue and its first $1 billion month in July, Friar told CNBC that the shortage of GPUs and compute remains the main bottleneck, which is why OpenAI is investing in mega data centers like Stargate. The company is diversifying partnerships with Oracle and CoreWeave, though Microsoft remains central to its tech and IP. OpenAI, valued at up to $500 billion, expects demand to keep climbing, with CEO Sam Altman saying the company could spend trillions on data centers.

Google has launched the Pixel 10 lineup at its Made by Google event, introducing three models (Pixel 10, 10 Pro, and 10 Pro XL) priced from $799 to $1,199. Powered by the new Tensor G5 chip, the phones bring over 20 AI features, including conversational photo editing, real-time phone call translation in 10 languages, and “Magic Cue” suggestions across Gmail, Calendar, and Messages. A new Visual Guidance upgrade lets Gemini Live provide on-screen cues, while a built-in Gemini Nano model boosts privacy with on-device AI. Google framed the launch as a major step ahead of Apple in AI-powered smartphones.

Google Photos is rolling out a new Gemini AI-powered editing feature that lets users change photos with simple voice or text commands. Instead of manually using tools, you can just say things like “remove cars” or “brighten colors” and the AI makes the edits. You can also chain multiple edits together or swap backgrounds entirely. On top of that, the new Pixel 10 adds C2PA content credentials, a visible label showing whether and how a photo was altered, giving viewers clear proof of AI edits.

Illinois has become the third U.S. state to ban the use of AI chatbots in mental health treatment, joining Utah and Nevada. The new “Therapy Resources Oversight” law forbids licensed therapists from using AI to make treatment decisions or communicate directly with patients, and it blocks companies from marketing chatbots as full therapy tools without human oversight. Lawmakers cite safety, privacy, and the risk of harmful guidance, with violations carrying penalties up to $10,000. The move follows studies showing chatbots sometimes encouraged dangerous behavior instead of offering proper help.

Two former Harvard students have launched Halo X, a $249 pair of AI-powered smart glasses that record, transcribe, and display real-time info during conversations. The founders say the glasses give users “infinite memory,” popping up answers like definitions or math results instantly. Backed by $1M in funding, Halo X runs on Google’s Gemini and Perplexity but relies on a phone app for computing. Unlike Meta’s smart glasses, Halo X has no external indicator for recording, sparking privacy concerns over covert use. Preorders open this week, though critics warn about consent and data security.

Meta has paused hiring in its AI division after aggressively recruiting more than 50 researchers and engineers from rivals like OpenAI, Google, and Anthropic. The freeze, confirmed by Meta, is part of a restructuring that splits its AI efforts into four groups, including a new “superintelligence” lab. The move comes after big spending on talent, with offers reaching hundreds of millions of dollars, and rising investor concern over stock-based pay. Analysts warn Meta’s lavish AI push could deliver major breakthroughs, or erode shareholder value if results fall short.

MOVIEFLO AI

MovieFlo.AI is where AI video meets professional storytelling: one platform for studios, marketers, and creators who think in scenes, not prompts.

In one smooth workflow, you can go from idea to final video with built-in tools that help you cast actors, plan scenes, and generate cinematic, consistent results.

Top video models like Veo, Kling, Wan, and Runway are right at your fingertips: no setup, no guesswork.

Built by veterans from LucasFilm and ILM, MovieFlo.AI delivers professional-grade tools with consistency baked in: from actors and props to branding and animation.

From product ads to short films, this is your production studio in a browser.

🎬 Try it for free, with 10% off annual subscriptions, exclusively for AI Valley readers → https://go.moviefloai.com/aivalley

*This is sponsored

TRENDING TOOLS

  • DeepSeek V3.1 - Updated model with smarter reasoning and a larger context window

  • April - Connects to Gmail and Google Calendar so you can ask hands-free questions while driving, like “What’s my first meeting today?”

  • Halo - $249 always-on AI glasses, built for everyday use

  • everyday.new - Personal AI that takes care of simple tasks and multi-step workflows

THINK PIECES / BRAIN BOOST

THE VALLEY GEMS

What’s trending on social today:

1/ Here is a super comprehensive list of coding agents.

2/ New Boston Dynamics demo is here.

3/ Elon Musk predicts AI will one-shot the human limbic system but Grok will increase the birth rate.

4/ GPT-5 invented new mathematics on its own.

THAT’S ALL FOR TODAY

Thank you for reading today’s edition. That’s all for today’s issue.

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

👍️ New reader? 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”.