did you know that the Amp team now publishes the daily mix of reasoning modes used by everyone (within Amp Code) and by themselves! the chart breaks usage into Low, Medium, High and Ultra, so you can see how often people actually reach for more compute.

it is a small but useful transparency move. the page shows percentages rather than request counts, so read it as a pattern, not a volume benchmark.

🔗 https://ampcode.com/models

fal has released 3DREAL Strong V2, an LTX-2.3 LoRA that turns rough 3D renders into photoreal video while trying to keep the original camera, layout and timing. the weights and a hosted endpoint are available.

the 3d render is really sick - watch the video; to be clear, they rendered the full video above with the 3D render below (from within blender)

🔗 https://huggingface.co/fal/LTX-2.3-3DREAL-LoRA

Google DeepMind CEO Demis Hassabis says human-level AI could be only a few years away. he wants the US to create an independent referee for the most powerful AI models, funded by the industry but overseen by the government. before a model launches, it would be tested for risks such as cyberattacks, biological misuse and bypassing its safety controls.

the idea is that every leading AI company should face the same checks, so nobody can cut corners just to launch first. if a model looks too dangerous, the referee could eventually delay its release or ask the whole industry to slow down. the hard part is making sure the AI companies funding the system do not end up controlling it.

we're inching ever so closer

🔗 https://x.com/demishassabis/status/2076957440109625718

vorflux launched as an “autopilot” for software engineering from former Rippling co-founder and CTO Prasanna Sankar. the company says it has raised a $15m seed to move ai coding beyond copilots that still need constant supervision.

the platform runs your full stack on dedicated cloud machines, splits work across agents and models for planning, building and review, then tests the result in a real browser before opening a pr. the bigger bet is that the valuable layer is no longer code generation, but the system that can reliably take work from idea to merge.

🔗 https://x.com/myprasanna/status/2077069901546852688

Notion can now open Markdown files straight from Finder as formatted, read-only previews, then turn them into editable Notion pages. the desktop app supports up to 10 files at once, which makes reviewing READMEs and notes a lot less awkward.

standard Markdown is supported, but anchor links and tool-specific extensions may need cleanup after import. like freaking finally 💀💀they're really late but ok props to them!

🔗 https://www.notion.com/help/import-data-into-notion

Mint released an MCP server and an open-source Three.js skill pack that let coding agents request 3D assets, pull the finished files into a project and assemble interactive apps or games around them.

the asset generation happens through Mint's remote, credit-based service, while the GitHub repo supplies the agent workflow and Three.js scaffolding. it is a useful pipeline, not a local 3D engine.

this is a project from an indie dev! crazy!!!!!!!!

🔗 https://github.com/mintdotgg/mint-threejs-skills

OpenAI's GPT-5.6 Sol prompting guide is basically an argument for smaller prompt contracts: define the outcome, constraints, evidence, success bar, and stopping conditions, then let the model choose the route.

OpenAI says leaner configs improved scores by roughly 10 to 15% in a sample of internal coding-agent evals while cutting tokens 41 to 66%, but those ranges are directional and should be tested on your own workload.

PS: just ask your clanker/agent to set it up for you! (on codex/or when using 5.6 sol)

🔗 https://developers.openai.com/api/docs/guides/prompt-guidance-gpt-5p6

something is coming for the construction industry!

Monumental raised a $32m Series B led by Khosla Ventures to scale its autonomous construction fleet. it already has 100+ robots laying bricks on real sites across Europe, with work completed for 100+ homes, a school, hotel, community centre and Amsterdam canal walls.

the interesting bit is the business model: contractors hire Monumental for the finished work, not the machines. its Atrium software turns architectural drawings into build plans and coordinates the robots with millimetre accuracy. the funding will expand the fleet, move beyond bricklaying and launch Monumental in the US this year.

🔗 https://www.monumental.co/press/announcing-our-32-million-fundraise

Thinking Machines (mira murati's startup; ex CTO @ OpenAI) released Inkling, a 975B-parameter mixture-of-experts model with 41B active parameters. it was trained from scratch on 45 trillion text, image, audio, and video tokens, supports up to 1 million tokens of context, and ships with the full weights.

the interesting angle is customization: Inkling is available for fine-tuning through Tinker, with native multimodal reasoning and a dial for trading thinking effort against cost.

🔗 https://thinkingmachines.ai/news/introducing-inkling/

good morning Kimi!

Moonshot AI launched Kimi K3, a huge 2.8T-parameter multimodal model with a 1M-token context window, built for long-running coding and agent work. it's live now across Kimi, Kimi Work, Kimi Code and the API.

the interesting bit is that it's being positioned as an open-weight frontier model, but the weights aren't actually out yet. Moonshot says they'll land by 27 july, with the full technical report still to come.

🔗 https://www.kimi.com/blog/kimi-k3

tldraw turned its whiteboard into a local desktop file that both you and coding agents can work on.

everything lives inside a portable .tldraw file, including the canvas, images, videos and reusable scripts. Codex or Claude Code can inspect the open board, create and rearrange shapes, or add new behaviour. no account needed, and it works offline. this feels less like a whiteboard app and more like a visual workspace for humans and agents.

🔗 https://offline.tldraw.com/

Decart's Lucy 2.5 can edit live video while it's happening. you can swap characters, add or remove objects, change backgrounds and styles, or generate effects from a prompt.

the interesting bit is what this unlocks beyond creator filters: virtual try-ons during live shopping, audience-controlled streams and product placement that changes on the fly. the public demo and api are available now.

🔗 https://x.com/DecartAI/status/2077801728213156044

Tencent Robotics X is teaching a home robot to give a massage while controlling both movement and pressure. the demo shows it reproducing several techniques, with the system tracking where the arms move and how much force they apply.

the interesting bit isn’t the massage. it’s a simple example of why robots working around people need touch and force control, not just cameras and a good-looking motion demo.

GET ME A TENCENT ROBOT NOW I NEED MASSAGES

🔗 https://x.com/XRoboHub/status/2078368180268102045

Maingen built SolarBench, an AI agent benchmark that puts models behind a simulated solar operations desk for a week. agents have to sort conflicting alarms, dispatch technicians, order parts and protect the portfolio’s P&L across 14 sites.

across eight tasks and 880 runs, Claude Fable 5 passed 53.8% of weeks. require four independent runs of the same task to all succeed and that drops to 23%. the bigger point is the failure mode: models often chased loud but cheap problems while missing quiet expensive ones.

🔗 https://solarbench.maingen.ai

Xiaomi has previewed Robotics-1, a vision-language-action model trained on more than 100,000 hours of wearable-captured manipulation data across 1,700+ scenarios.

the demo spans household tasks, but the evaluation needs careful framing: four tasks were tested out of the box, while laundry loading and packing used task-specific fine-tuning. Xiaomi has not released the Robotics-1 code or weights yet.

🔗 https://robotics.xiaomi.com/xiaomi-robotics-1.html

withings’ BodyScan 2 turns a bathroom scale into a 30-second daily check-in and a deeper 90-second weekly health scan. it tracks 60+ signals including six-zone body composition, vascular age, blood oxygen, nerve response and a six-lead ECG, with the results shown on a screen built into the handle.

it costs $599.95, and the fine print matters: some insights require Withings+, features are still rolling out, and most readings are wellness estimates rather than medical diagnoses.

would you buy one? definitely better than that xiaomi one most asians have 😭

🔗 https://www.withings.com/us/en/landing/bodyscan-2

Google's new Gemini lineup splits the fast tier three ways: 3.6 Flash is the all-round workhorse, with better coding, computer use and knowledge work. Google says it used 17% fewer output tokens than 3.5 Flash on Artificial Analysis, while 3.5 Flash-Lite clocked 350 output tokens per second at $0.30/$2.50 per million input/output tokens.

there's also a security-tuned 3.5 Flash Cyber for finding and patching vulnerabilities, but it's limited to governments and trusted partners through CodeMender. 3.6 Flash and Flash-Lite are available today; Google also says 3.5 Pro is in partner testing and Gemini 4 pre-training has started.

🔗 https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber/

Greptile found that AI models were slightly better at reviewing code written by a rival model than their own. across 1,000 PRs and roughly 1,500 verified bug comments, GPT caught more serious bugs in Claude-generated code, while Claude did better on Codex-generated code.

its experimental “model inversion” feature detects which coding agent likely wrote a PR from commit trails, branch names and titles, then routes the review to the other model. a useful reminder that a second model can expose different blind spots.

🔗 https://www.greptile.com/blog/model-inversion

Block has released Buzz, an open-source group chat for teams where people and AI agents work in the same rooms with the same project context. agents can join channels, work with git projects, review code, run workflows and collaborate with each other instead of living behind separate one-shot prompts.

it’s built on Nostr, model-agnostic and self-hostable, with desktop apps for macOS, Windows and Linux. it’s still early though, with mobile apps, approval gates and some huddle features still being wired up.

🔗 https://buzz.xyz/

Beyond the Prompt 2.0 is bringing Singapore’s product and design community together for a practical look at how people are actually building with AI, beyond basic prompting.

expect talks on shipping an iOS game, building design taste without a design background, and scaling AI-generated motion across products and teams. it’s free, but registration needs host approval. happening 29 july, 6:30pm in Singapore, with the exact venue shared after registration.

come join if you're free! 🫶

🔗 https://luma.com/j8ob174c