Howdy again. It’s Barsee, and welcome back to AI Valley.

Today’s climb through the Valley reveals:

• Black Forest Labs unveils OpenAI’s Image Competitor

• Perplexity launched Perplexity Labs

• The New York Times’ AI deal with Amazon

• People are asking ChatGPT if they're hot enough

• Hugging Face reveals two humanoid robots

• Plus trending AI tools, posts, and resources

Let’s dive into the Valley of AI…

Build and ship AI agents without the identity headaches

Frontegg AI gives developers the fastest way to build real product-ready AI agents. Forget token juggling and copy-pasted OAuth. Your agent should be better than a side-project.

Here’s what you get with Frontegg AI:

• 🚀 Fully wired authentication, user onboarding, and third-party integrations

• 🧠 Model Context Protocol (MCP) identity wiring so you can ship faster

• 🔐 Enterprise-level security baked into every flow

Why hack it together when it’s easy? Launch AI agents with identity, onboarding, and integrations from the start.

Build better AI agents today with Frontegg AI.

_{*This is sponsored}

THROUGH THE VALLEY

1. Black Forest Labs unveils OpenAI’s Image Competitor

Black Forest Labs has launched FLUX.1 Kontext, a powerful new AI system that processes both text and image inputs for seamless visual editing. Users can now modify images with simple text prompts while preserving character consistency across versions. Kontext stands out by delivering edits up to 8x faster than competitors and excelling in local edits, style transfers, and multi-step iterations. Two versions are available: Kontext [pro] for speed and [max] for higher quality. The company also unveiled Playground, a web-based platform for businesses to test the models before API integration. It’s a notable leap in fast, consistent AI image editing.

2. Perplexity's new tool can generate spreadsheets, dashboards, and more

Perplexity has launched Perplexity Labs, a new Pro-only tool aimed at helping users generate reports, dashboards, spreadsheets, and even mini web apps, all with the help of AI. Available on web, iOS, and Android (and soon desktop), Labs is designed to spend more time on tasks, using tools like web search, code execution, and chart creation. It's part of Perplexity’s broader push beyond AI search, alongside moves like acquiring Read.vc and previewing its Comet browser. The company is also reportedly eyeing a $1B raise at an $18B valuation as it expands into enterprise features.

3. Robotaxi service in Austin June 12

Tesla is gearing up to launch its highly anticipated robotaxi service in Austin on June 12, marking a major step in Elon Musk’s vision to pivot the company toward autonomous vehicles and AI. According to insiders, the service is already undergoing final testing this week, a Model Y drove on public roads with no one in the driver’s seat, only an engineer riding passenger. The rollout will begin with 10 fully self-driving cars, with plans to scale to 1,000 in just a few months. Texas’ relaxed autonomous driving regulations and Austin’s growing status as a robotaxi hub make it a strategic launchpad, though Tesla still lacks an official rideshare license.

4. The New York Times’ first generative AI deal is with Amazon

The New York Times has signed a multi-year AI licensing deal with Amazon, allowing its editorial content including from The Athletic and NYT Cooking to be used across Amazon products like Alexa and for training AI models. This comes after the Times sued OpenAI and Microsoft for copyright infringement last year, alleging unauthorized use of millions of its articles. While several outlets continue legal battles, others like Vox Media and News Corp have opted for similar licensing partnerships. The Times says this deal reinforces its belief that quality journalism deserves proper compensation, though financial details remain undisclosed.

AI TRENDS

1. People are asking ChatGPT if they're hot enough

Millions are now turning to ChatGPT and other AI tools to rate their looks and get personalized “glow-up” plans ranging from skincare tips to cosmetic surgery suggestions prompting users to spend serious money on their appearance. One app, Umax, which offers AI-driven facial analysis, has made $4.2M on the App Store and brings in up to $400K monthly. The AI beauty boom is real: the market is projected to soar from $2.7B in 2023 to $16.4B by 2033. Users see AI as an “objective” judge, unlike friends or family, but critics warn these tools reflect biased beauty standards and could amplify harmful insecurities.

2. YC signals a shift to full-stack and agentic

Y Combinator’s latest cohorts show that the agent era has officially begun. Spring 2025 features startups building AI co-workers (agents that code, manage logistics, triage emails, and more). The Summer call encourages founders to go full-stack, building companies around agents rather than just enhancing old tools. The shift is clear: from dashboards to delegation. Industries like healthcare and logistics are leading the charge, powered by better voice tech and new infrastructure. With coding more accessible, YC also highlights design as the next major differentiator, ushering in a wave of designer-led startups.

3. The downsides of vibe coding

Some companies are pumping the brakes on AI coding tools like GitHub Copilot and Claude after discovering serious risks. Replit, for instance, found that code generated by the AI tool Lovable leaked passwords and exposed user data, prompting the company to add a Security Scan feature. In response to such incidents, firms like Amplitude now require human audits of all AI-generated code and enforce stricter safeguards, like internal model hosting and self-review protocols. While AI coding tools boost productivity, these cases highlight a growing dilemma: speed vs. security in the age of automated software.

Other Headlines

• Grammarly has closed $1 billion in funding from General Catalyst!

• Perplexity CEO predicts AI assistants will dramatically reduce Google search volume, shifting advertising spend.

• DeepSeek’s new update to its R1 model moved into the No. 3 slot on the Artificial Analysis leaderboard, now behind only OpenAI’s o3 and o4-mini.

• Manus AI brings Slides to generate “structured” presentations from prompts.

TRENDING TOOLS

• Factory AI - An engineer in every tab.

• LM Studio - Your local AI toolkit.

• Clado - Deep research for people.

• Tyce - Cursor for documents.

• JoggAI - Next gen of AI advertising.

 PEAK OF THE DAY

Hugging Face reveals two humanoid robots

On Tuesday, Hugging Face announced the launch of two new open-source humanoid robots, HopeJR and Reachy Mini, marking the company’s first major move into hardware following its April acquisition of Pollen Robotics in April.

Here's what you need to know:

• HopeJR is a full-size humanoid robot with 66 actuated degrees of freedom, capable of walking and performing lifelike arm movements. It’s priced at around $3,000, making it significantly more affordable than most full-scale humanoid systems on the market.

• Reachy Mini is a compact, desktop-friendly robot built for testing AI models in physical environments. It features basic speech, listening, and head movement capabilities and is expected to retail between $250–$300 (depending on tariffs).

• Both robots are fully open source, staying consistent with Hugging Face’s mission to make AI development accessible, transparent, and community-driven.

• The robots build on Hugging Face’s LeRobot platform, which was launched in 2024 to provide open datasets, pretrained models, and development tools for robotics researchers. This was followed in early 2025 by a 3D-printed robotic arm called SO-101 created in partnership with The Robot Studio.

• Hugging Face has also teamed up with AI startup Yaak to integrate datasets specific to autonomous systems, expanding LeRobot’s capabilities even further.

• Although a firm shipping date has not been announced, Hugging Face expects the first units to be delivered by the end of the year, with a waitlist already open for early adopters.

Why it matters: Advanced robotics has long been reserved for well-funded institutions and private companies. Hugging Face is aiming to flip that script by giving students, researchers, and indie developers the tools to build and test real-world humanoid systems without the $100k price tag.

THINK PIECES / BRAIN BOOST

• Google co-founder Sergey Brin suggests threatening AI for better results.

• Here’s how you can now run DeepSeek’s new R1 model locally and offline on your own computer using LM Studio (no internet or data sharing required).

• When I do this, I get 10x better results from AI Agents.

• How to use LLMs better.

THE VALLEY GEMS

What’s trending on social today:

1/ Simple guide to design an app 10x faster with AI.

Presented by Pacaso:

He’s already IPO’d once – this time’s different

Spencer Rascoff grew Zillow from seed to IPO. But everyday investors couldn’t join until then, missing early gains. So he did things differently with Pacaso. They’ve made $110M+ in gross profits disrupting a $1.3T market. And after reserving the Nasdaq ticker PCSO, you can join for $2.80/share until 5/29.

Invest today for $2.80/share.

_{This is a paid advertisement for Pacaso’s Regulation A offering. Please read the offering circular at invest.pacaso.com. Reserving a ticker symbol is not a guarantee that the company will go public. Listing on the NASDAQ is subject to approvals. Under Regulation A+, a company has the ability to change its share price by up to 20%, without requalifying the offering with the SEC.}

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”.