Sponsored by
|
BUILDER HUSTLE
AI · TECH · STARTUPS
|
|
|
In This Issue
|
|
🤖 Gemini 3.5 Flash Can Now Control Your Screen
|
|
📚 Google's Free Study Notebooks Are Now Live
|
|
📱 Someone Vibe Coded an iPhone Into a Click Wheel iPod
|
|
🔒 OpenAI May Have Quietly Launched GPT-5.6
|
|
|
🤖 GEMINI 3.5 FLASH · COMPUTER USE · AGENTS
|
|
Gemini 3.5 Flash Can Now See and Control Your Screen.
|
|
|
Google baked Computer Use directly into Gemini 3.5 Flash as a native built-in tool. Developers can now build agents that see screens and take real actions across browser, mobile, and desktop environments without wiring up a separate model. It unlocks long-horizon automation like continuous software testing, knowledge work across professional apps, and cross-platform workflows. It is live today via the Gemini API and the Gemini Enterprise Agent Platform, with safety rails that require explicit user confirmation before any sensitive or irreversible action is taken.
|
|
📌 Computer Use is now native to 3.5 Flash, not a separate model. One API call gets you screen vision, reasoning, and action all in one.
|
|
⚡ 3.5 Flash already beats Gemini 3.1 Pro on coding and agentic benchmarks and runs at roughly 4x the token output speed of comparable frontier models.
|
|
🔭 Automated task stopping kicks in if threats are detected mid-run, which makes this actually usable in enterprise without a human babysitting every step.
|
|
|
Read the Google announcement →
|
|
The best voice models, now across all channels
Most CX platforms do not own the voice. They orchestrate a workflow, then call a third party for speech and transcription. Every hop adds latency, cost, and another vendor to manage.
ElevenAgents is the opposite. They make the voice models the market builds on, and ElevenAgents puts full orchestration on top. Voice, transcription, text-based chat, and reasoning run in one vertically integrated pipeline, so responses come back in <400 milliseconds and sound human, not synthetic.
Plus, you keep full control. Plug in any LLM, integrate tools, webhooks, and MCP servers, and ground responses in your knowledge base. Get an agent live in minutes, then A/B test with Experiments, enforce Guardrails, and version every change.
The payoff: more human conversations, lower latency, and far less time stitching infrastructure together. You build on the models you already trust. Pricing is transparent and flat at $0.08 per minute.
|
📱 IPHONE · DUMBPHONE · VIBE CODE
|
|
Someone Vibe Coded an iPhone Into a Click Wheel iPod.
|
|
A developer built a full iPhone app and 3D-printed case combo that physically turns your iPhone into a retro iPod-style dumbphone. The case covers most of the screen with a real working click wheel. The custom OS underneath has 15 focused mini-apps including music with Cover Flow, books, a camera with film LUTs, notes, voice memos, and fitness tracking. Swappable retro themes, no doom-scrolling, no notification spiral.
|
|
📌 The click wheel is a physical case, not a screen UI. It literally blocks the display so the temptation to scroll is removed at the hardware level.
|
|
⚡ 15 mini-apps cover the essentials - music, camera, books, notes.. without any social feed or browser in sight.
|
|
|
See the full build thread →
|
|
|
🔒 OPENAI · GPT-5.6 · ALLEGED
|
|
OpenAI May Have Quietly Launched GPT-5.6 for a Few Partners.
|
|
|
Early reports claim OpenAI has launched a new family called GPT-5.6 with three tiers: Sol (flagship), Terra (half the cost of current top models with similar performance), and Luna (fastest and cheapest for high-volume work). The rollout is reportedly limited to trusted partners at US government request, with Sol allegedly matching Mythos on ExploitBench while using roughly a third of the compute.
|
|
📌 Three-tier naming - Sol, Terra, Luna.. would mirror the way Anthropic and Google have started segmenting by price and performance rather than version number alone.
|
|
⚡ Terra at half the cost of current top models with similar performance would be the most interesting tier for most builders, better value than anything OpenAI offers today.
|
|
|
Watch OpenAI for official news →
|
|
|
📚 GOOGLE GEMINI · STUDY NOTEBOOKS · FREE
|
|
Google Just Gave Every Student a Free AI Tutor.
|
|
Google launched Study Notebooks inside the Gemini app globally, no paid subscription needed. You upload your materials or pick a learning goal and Gemini builds a personalised study plan: diagnostic quizzes, bite-sized lessons, practice problems, and progress tracking with strengths and focus areas highlighted. It also plugs into NotebookLM for flashcards and video overviews. SAT prep is already supported with JEE, NEET, ACT, GRE, and ENEM on the way. It is live on web now, with mobile and school accounts rolling out soon.
|
|
📌 Fully free, globally available, every language Gemini supports.
|
|
⚡ Diagnostic quizzes adapt to your goal and uploaded content, so every student gets a different learning path based on what they actually need.
|
|
🔭 The NotebookLM integration means you can go from a lecture PDF to flashcards and an AI-generated video overview without leaving the Gemini app.
|
|
|
See Study Notebooks on Google →
|
|
|
Top Products To Watch Today
|
|
Open Source · Website Builder · AI
Free, local, open-source alternative to Lovable, v0, and Bolt. Build and run AI-generated apps entirely on your own machine.
|
|
|
Messaging · Social Media · AI
An AI-native Slack alternative built for big teams and communities. Smarter messaging without the noise.
|
|
|
Video · AI
HeyGen calls it the world's first Creative Operating System. Feed it your goals and it handles the full video production pipeline.
|
|
|
Stay sharp,
Builder Hustle
