sunsetroad

All the money in the world can't buy you a second of yesterday... Live the moment...

[sunsetroad @sunsetroad]

🙏♾️

…thought matters…

https://t.co/qlqXKp2yZW

[Z.ai @Zai_org]

GLM-5.3 is now open-weight.

Our most capable model for agentic coding and cyber defense is now available to download, run, and customize.

Weights: https://t.co/v1IbWMXxg4
Tech blog: https://t.co/ekQkO83jCv

[Susan Zhang @suchenzang]

i'm sorry to my overlords, this was funnier than expected

[Simon Willison @simonw]

My LLM cliché highlighter is up to 38 patterns now https://t.co/sew4qNNaZX

[eric zakariasson @ericzakariasson]

you can now share grok bots! here's the projects manager to manage teams of grok bots

1. install bot: https://t.co/QG7yo1GIMV
2. ask the bot to create specialists
- coder
- designer
- researcher
- writer
3. create your first project with projects manager!

[Runway @runwayml]

You've never seen anything like this.

[sunsetroad @sunsetroad]

@0xSero 😉🙌

[DAIR.AI @dair_ai]

Banger paper from Google.

If you maintain a skill library for your agents, you might want to check this out.

(bookmark it)

This work separates three things that skill-evolution systems usually collapse into one. Raw execution traces, a persistent wiki of accumulated knowledge, and the executable skills themselves.

Experience gets consolidated into the wiki, and every later skill update builds on that wiki instead of on a scattered optimization history.

Ablations confirm the wiki is what carries a lot of the gain. Two results stand out in particular. Smaller models with evolved skills beat substantially larger models without them. And skills evolved by one model transfer across families, where skills evolved elsewhere sometimes beat self-evolved ones.

Paper: https://t.co/6qftGirTpE

Chat with Paper: https://t.co/rrVzkkR1ij

[Shubham Saboo @Saboo_Shubham_]

Crazyy...someone measured Claude's vocabulary across 47,464 GitHub pull requests.

A word cluster that didn't exist in 2025 is now 45% of human-authored PRs.

The top word is "load-bearing"

[Shruti @heyshrutimishra]

Unitree's humanoid robots went viral doing kung fu at China's New Year Gala.

Search volume jumped 300%. Orders doubled within hours.

Then the company went public on Shanghai's STAR Market. The IPO raised $904 million. The stock surged 460% on day one, peaking at 629% above the offer price. Retail investors oversubscribed it 8,000 times.

China now ships 97% of the world's humanoid robots.

Unitree and AgiBot control three quarters of the market. Global shipments surged 272% YoY in the first half of 2026.

The stock has since plunged 45% from its peak.

[Google Research @GoogleResearch]

Introducing the planetary prediction engine: An experimental research capability that automates global geospatial modeling workflows. From predicting public health outcomes to improving food security, we are simplifying planetary-scale analytics. Learn more: https://t.co/EGhcEMqZxX

[Alex Finn @AlexFinn]

Grok Bot is by far the best AI agent out there right now

The issue is a lot of people just have no idea what to do with it

Here is everything my Grok Bot got done for me this week (use these as inspiration):

1. Autonomously found a challenge someone was having on Reddit and built a microSaaS to solve it

2. Built an entire 3D digital office so I can watch all my bots work on a 2nd monitor

3. Walked me through installing and setting up Omarchy on my old gaming computer

4. Built a journaling TUI plugin for Omarchy so I can hit a keybind and quickly make an entry

5. Initiated and completed an entire return process for some trendy T shirts I bought that I'm way too old to wear

6. Has been collecting AI news stories every day then posting them in my community

7. Watching bug feedback on my app Creator Buddy and autonomously fixing them and pushing the PRs using cursor cloud agents

8. Checking X every 5 minutes for new model releases so I don't have to doom scroll as much

9. Added a close button to notifications in Omarchy that don't by default have them

10. Negotiated 2 brand deals and managed the entire process for me

11. Read 2 contracts and sent red lines back for other contracts I'm negotiating

The key to finding good use cases for Grok Bot I find is anything you are about to do on your computer, first ask Grok Bot if it can do it first

You'll actually be shocked by how much it can take off your plate

[BOOTOSHI 👑 @KingBootoshi]

HOLY FUCK IT WORKS

MY MAC MINI IS CALLING ME PRIVATELY WITH FACETIME AUDIO

HAHAHAHAHAHAHAHAHAHAH

[Dyna Robotics @DynaRobotics]

After so many demos, models, pilots, robots are still struggling to land real deployments with real customers. Until now.

Today we’re excited to share that our robots have successfully crossed the ROI threshold, and Din Tai Fung, one of the highest revenue per location restaurant chain in the US, is rolling out Dyna robots across its extensive restaurant network.

This brings our rollouts across hotels, logistics, data centers, and many other use cases to a fleet that reaches hundreds of robots by the first half of 2027. And we’re just getting started.

It’s been a wild year, and today we’re double clicking on the battlefield stories and sharing a few learnings about scaling robot deployments. We are just scratching the surface.

Read the full blog post: https://t.co/8QVEoQuzBY

[Marc Benioff @Benioff]

Welcome Claudeforce. ❤️

[ChatGPT @ChatGPT]

ChatGPT Work can now use its computer and browser to sign in to websites on web and mobile, without ChatGPT ever seeing your username or password.

That means you can ask it to:
• Set up utilities for a new apartment
• Book a DMV or passport appointment
• Check reimbursement costs through your insurance
• Find and book an in-network doctor around your availability
• Check when your car registration expires and prepare the renewal paperwork
• Compare your rental insurance policy with an issue you’re emailing your landlord about
• Find and save apartment listings that match your criteria
• Restock something just by uploading a photo
• Schedule a package pickup for a return
• Cancel tickets for a rescheduled trip
• Book a vet appointment
• Submit reimbursement paperwork for medical treatments
• Check resale sites for new drops and save things you might like
• Find candidates with specific experience and draft outreach
• Take invoices from your email and submit them to your accounting software
• Draft replies to rental property inquiries
• Fill out permit applications for your small business
• Add action items to a vendor portal based on a recent client call
• Analyze the latest ad campaign for your small business

[Lou @louszbd]

We use GLM-5.3-Flash build a dream kitchen. A 3D world built in Blender. This is not a generated video.

[DAN KOE @thedankoe]

https://t.co/klNSdaI7aa

[Ethan Mollick @emollick]

A line in AI video was crossed, in my experiments with just the web interface, H3 Max can now create reasonably high quality AI video in less time than it takes you to watch it. This is realtime from the moment I pushed the "generate" button (and also includes prompt enhancement)

[Kun Chen @kunchenguid]

i’ve been building and using agents since the gpt-4 days

back then, the model couldn’t even reliably edit files. it couldn’t generate a valid unified diff, couldn’t preserve white spaces in find/replace tool calls, plus various kinds of quirks

we implemented many harness level tricks to compensate, it got a lot better than using the raw model, but it still sucked

then sonnet 3.5 v2 came out and it was trained with a coding harness with RL. it knew how to find/replace reliably. it knew how to compose good bash commands to get what it needs. all the tricks we did were reverted

that was the first time i internalized the bitter lesson https://t.co/Cy9uGNdbWy

most of what we are doing in the harness today, all the smart context engineering tricks, all the useful markdown files we throw into our repo, they will all go away

eventually, the models just know

there will be a model that will do incredibly efficient and smart compaction that preserves what the continued session actually needs

there will be a model that know how to plan with you better than any skill you can find today

there will be a model that will do perfect code reviews with just the right feedback

the models just know

i believe as humans today, most of us should not be fiddling too much with harness tricks. if you don’t believe they will go away, you can surely still believe they will keep changing rapidly to a point where anything you learn becomes obsolete every several weeks

instead, focus on what i call “the 3 fundamentals” -

1. understanding the world

this is your input. how does the world work? what are people doing? what problems do they have?

knowing how to ground yourself with real world knowledge helps you avoid hallucinations, and work on things that actually matter

2. first-principles thinking

this is your compute. with the real world understanding you gathered, what insights can you derive that’s likely to be true? what predictions can you make about the future?

being able to think in a disciplined way is what allows you to arrive at useful conclusions that can guide your actions

3. articulating our thoughts

this is your output, and it’s your AI’s input. it may sound easy but it’s a real skill. not everyone can communicate effectively, whether it’s to humans or AI

with AI eventually becoming incredibly capable, the clarity in the articulation of our intent is the main, if not the only, bottleneck

these 3 fundamentals do not shift as models improve. they only become more and more critical as the bitter lesson manifests

for the vast majority, i suggest spending your time on what will still matter in a year, five years, a decade. let the geeks play with what’s hot this week - in the end, you will see that you didn’t miss anything

[Aporia @aporia9n]

Really believe that ADHDs, if they play their cards well, might be the biggest winners of this AI gold rush

Cause unlike corporate careers rewarding consistency & compounding, the new world disproportionally rewards curiosity, obsession, rapid switching and intensity

The basic problem with ADHD in the old world was that curiosity was expensive. You’d get obsessed with some random idea, spend 3 days learning everything about it, want to build something, then realise you need 6 months to learn how to code / design / whatever (or pay someone who can) and eventually move onto the next thing

But AI completely changes that equation because the distance between “wait this would be cool” and actually testing it has collapsed from months to sometimes hours. You can get curious about some niche industry in the morning, understand the basics by lunch, vibe code something in the afternoon and potentially have people testing it that evening. So this disproportionately rewards people whose natural state is basically jumping between rabbit holes, connecting unrelated things and occasionally becoming insanely obsessed with one of them

AI is also slowly deleting the boring middle. A lot of ADHD people aren’t necessarily bad at starting things or going insanely deep when something captures their attention, they’re bad at the 200 repetitive steps between “this is exciting” and “this is finished” (boilerplate, formatting, basic debugging, admin, iteration, etc.) And those are increasingly exactly the things you can delegate to AI, which means the economic value of being insanely curious + occasionally obsessive probably goes up, while the cost of being bad at repetitive execution goes down

But there’s a massive catch, because AI is also creating the greatest distraction machine ever invented. Every morning there’s a new model, agent, tool, business model or “once in a generation opportunity” to chase, and the same brain that benefits most from cheap experimentation is also the one most likely to spend 5 years being extremely early to everything and compounding absolutely nothing

So the actual winning combination might actually be ADHD-style exploration + artificially imposed constraints: explore aggressively, follow weird rabbit holes, test way more things than a normal person would, but once something works force yourself into sprints where you’re simply not allowed to play another game

For most of history the economy disproportionately rewarded people who could reliably do the same useful thing for 8 hours a day, 250 days a year. For this weird period of AI, the returns temporarily shift towards people who can obsess over 20 things, connect 2 that nobody else connected, then disappear for 3 days and build the thing