How having no shame can get you to $500K/month

🎤 Voice as infrastructure – Play AI offers hyper-realistic voice generation across multiple languages, accents, and dialects. Its technology enables natural-sounding, customizable voices designed for integration into AI-powered applications, including customer service, content creation, and real-time assistants.

🧠 Strategic fit across AI product lines – According to an internal Meta memo, Play AI’s work will support several initiatives: Meta AI’s core assistant, AI Characters for interactive use cases, and future wearable experiences. The acquisition aligns with Meta’s broader efforts to differentiate its AI offerings through more immersive, multimodal interaction.

💸 Integrated into a wider hiring and acquisition strategy – Play AI had previously raised $23.5 million and partnered with Groq and Reality Defender on both performance optimization and voice deepfake detection. Its acquisition follows a broader wave of hiring at Meta, which has recently recruited talent from OpenAI, Apple, Scale AI, and SSI. Signing bonuses for key researchers have reportedly reached up to $100 million.

🕶️ Supporting Meta’s wearable ambitions – This acquisition complements Meta’s $3.5 billion investment in eyewear company EssilorLuxottica and its growing focus on AI-powered wearables. Combining voice, vision, and embodied AI tools suggests Meta is investing in a fully integrated hardware-software platform.

🛠️ Start as a service, not a startup – Ben recommends launching with a focused AI automation agency to get paid, learn fast, and build the systems people already want, before even thinking about SaaS.

🔄 Build one system, not ten – Find one high-ROI automation in a niche (e.g. lead gen for dentists) and turn it into a repeatable product. Less custom work, more margin, easier to sell.

💰 Cash flow funds the jump – Product takes time. Building services first lets you fund the pivot without outside money or betting the house.

📈 Lead flow and client fit are non-negotiable – Ben breaks down how predictable sales, tight qualification, and long-term retainers give you the foundation to scale up without spinning out.

🎯 The right model depends on you – Want fast cash? Go the agency route. Want to build a product? Prove it first. Either way, know your strengths and goals, don’t just chase what’s trending on X that day.

🔍 Lovable pages are invisible by default – Client-side rendering means your H1s, keywords, and body text don’t show up to Google. Without pre-rendered HTML, your site basically doesn’t exist to search engines.

🛠️ The workaround: static site generation – Using a custom “magic prompt,” you can instruct Lovable to export HTML pages instead of rendering them on the client. That makes your content crawlable—and unlocks real SEO.

⚠️ Blog posts break without an extra step – The static export only covers top-level pages by default. Without a second prompt to handle nested folders like /blog, posts won’t get indexed. A small change with big impact.

🚀 Netlify fixes the publishing issues – Hosting on Netlify (instead of directly via Lovable) ensures all your static pages deploy correctly, including blogs and subfolders. It also supports custom build commands for precise control.

📈 The right build command matters – Changing Netlify’s build command from build client to bun run build ensures the static export works properly. A minor tweak, but critical for full SEO coverage.

🧠 An AI that helps without being asked – Cluey runs in the background, seeing and hearing what you do on-screen, and steps in with relevant help in real time. No prompts, just passive support when you need it.

🔥 He started with a tool that broke the rules – His first idea helped people cheat tech interviews. Instead of hiding it, he shared it publicly, and it went viral. That traction helped him raise money and build Cluey.

📱 Short-form content is the whole strategy – He works with a bunch of creators who post constant short videos across TikTok, X, and Instagram. The idea is volume and velocity, not perfectly curated campaigns.

😈 Roy runs the company account himself – His personal voice is the brand. He posts hot takes, shares behind-the-scenes updates, and isn’t trying to sound like a startup CEO.

💡 The endgame is to own distribution – The big bet isn’t just the tech—it’s that whoever builds the AI layer people actually use day to day will win. Cluey’s aiming to be that layer before the big labs get there.

🧠 Spec-first, code-second – Kira turns prompts into detailed specs, user stories, and implementation plans before writing any code. This upfront planning brings clarity and aligns with real-world engineering workflows.

🧩 Hooks automate what devs forget – Event-based triggers update documentation, run tests, or refactor code automatically. They’re saved as files and can be versioned and shared across teams.

💻 Feels familiar, not foreign – As a VS Code fork, Kira supports Mac, Windows, and Linux, and lets users import their existing profiles and extensions—lowering the barrier to adoption. It’s also free while in preview, giving teams room to test its full capabilities before committing to a paid plan.

🛠️ Claude under the hood – Kira uses Anthropic’s Claude models to power both chat and spec modes, allowing developers to choose between "vibe" (chat-first) and "spec" (plan-first) workflows.