How I hacked my timeline, and the Chinese AI Culture:
Yesterday, a post from @dontbesilent appeared in my feed. He was wondering why foreigners who can't read chinese were suddenly answering to his tweets (that was me), and if we were using auto-translate.
I replied with the reality: long before X had the auto-translate feature enabled, I realized the chinese content was way better than the english feed. So, I started manually copying and translating their articles. I intentionally trained my algorithm to feed me their posts because the insights were just better (imo).
That one reply sparked a massive back and forth yesterday with the chinese community in the comments. The cultural exchange was incredible, and it perfectly mirrored what i saw on my visit to China this last christmas.
The Implementation Reality
I spent over two weeks moving through the country: Shenzhen, passing by Xingping, Phoenix, Tianmen, Zhangjiajie (prob the most beautiful place on earth), Xi'An, Shanghai, Beijing, Guangzhou, and back home.
Everyone in the west talks about ai, fintech and implementation, but China is years ahead. To give you an idea of the scale:
We were driving through a random, remote village somewhere between phoenix and zhangjiajie. Literally the middle of nowhere. We stopped at a crosswalk by a dirt roundabout, and there was a guy selling fruit in a cartwheel.
Even this remote dirt-road fruit vendor had an alipay & wechat qr code. We paid with our phones (and generously tipped him bc we couldn't believe he could be in that remote place selling fruit).
The entire country runs on a seamless digital infrastructure that makes the rest of the world look like its stuck in the past. There is no place on earth where there is such a general technology adoption.
The Friction & The Physical Comedy
My worst and only problem of the entire trip was the language barrier.
What I love the most every time I travel, is mixing with locals. Understanding their mental models, exchanging ideas, knowing their culture and ways of thinking. From the smallest detail in their day to day life, to the way of working of the entire society of the country I am in.
For the first time in my life, I couldn't do that. The friction was just too high (the younger generations are doing great though. Kids would constantly approach me just to practice a few english words).
Also, being redhead in rural china is a really interesting experience:
People would take sneaky photos every day, parents pointing me out to their kids, Chinese locals asking me to take photos with them, even a mother literally handed me her baby just to take a smiling photo with me.
The people are incredibly welcoming, but the communication gap is (very sadly), too big right now.
Thoughts after my trip
I loved it so much, I seriously looked into moving there for a few months right before I moved to Portugal.
Unfortunately, the visa requirements allowed me just to stay for 30 days as a tourist. Otherwise I would have to create a company, or get hired by one (none of them were options in my plan for now).
But here is the absolute reality: at their current pace and work ethic, they are going to completely dominate the tech and ai meta. I have no doubts about this.
I will definitely return at some point to live there for a few months and being able to dive really deep into AI, local models, hardware, and above all, the culture and people.
But next time I am bringing a real time ai auto translate device with me, so I can actually talk to everyone over there ;)
(and massive thanks to everyone on the chinese timeline who spent their day exchanging thoughts with me yesterday. the bridge is being built.)
I use Grok Bot to keep my home timeline worth reading. Every day Grok Bot reads my whole timeline, sorts high quality posts from low quality ones (engagement bait, unlabeled ads, recycled threads, spam), and saves the good ones to labeled bookmark folders for me to directly read. That part works well. It saves me time from reading so much bullshit going on lately on x.
What's missing is the part that matters most: training my feed. Grok Bot's X connection has no way to:
-mark a post as "Not interested in this post"
-mute an account (or a word or phrase)
-give other negative signals for the recommendation algorithm, like "show fewer posts from this account"
So once Grok Bot spots junk, all it can do is suggest blocking the whole account. That's a heavy tool for something that should just be feedback to the algorithm. Grok Bot also isn't allowed to use in a browser, even when I sign in myself, so there's no other route.
Request to @bot team to add these as tools in Grok Bot's x plugin:
-"Not interested" on a specific post
-Muting and unmuting accounts
-"Show fewer from this account"
Grok Bot could then clean up a feed the way users already can by hand, without anyone having to block accounts. And this would make our timeline high quality back again.
Idk why lately the algorithm just shows nonsense with mostly no real value
7
6
290
4. Don’t lose momentum
The best career advice for young people is to just focus relentlessly on 3 things:
1. Get skilled. Technical skills, writing skills, people-orchestration skills, everything.
2. Get involved contributing to a mission you're aligned with. It doesn't have to be pure bliss and your contribution doesn't have to match your top strengths exactly. But at the end of the day you have to feel some pride in what you did or what you learned. If you can find something like that while becoming skilled in a number of different ways then you'll find ways to increasingly wrap your skill profile around it (and you'll find/develop strengths in your skill profile that you never expected).
3. Pick up momentum. If you do (1) and (2) right then you've got direction and potential and all that's left is to sprint forwards and achieve it. The more momentum you pick up, the better you get, the more real value you produce, the more opportunities you create, the more you get noticed, the more value you capture into your own life, the more effort you want to put in... you create a virtuous cycle with so much momentum behind it that it becomes an unstoppable virtuous cyclone.
174
Just had GrokBot contact, negotiate, and schedule visits to get quotations with all available gardener companies in the area for my countryside house.
Something we never did because of laziness of finding the best option, now completely automated with @grok
@bot . Starting to see the potential of it.
Just create an email specific for GrokBot, give it complete access, and let it do things.
1
131
insane how every feature @levelsio posts about improving claude gets officially implemented in the next weeks
You can now set Claude Code's output style to Concise.
Claude leads with the result, keeps responses short, and still gives full detail when you ask.
Turn it on in /config → Output style, or set "outputStyle": "Concise" in settings.json.
1
179
This is a major hack i've been working with for the last weeks.
Added this one rule: every turn ends with these same four blocks.
What you need from me. What you did. My options. What happens if I say nothing.
My active work looks like this. I have between 5-6 active desktops, each for a different project/product. In each desktop, I run multiple terminals (claude,gpt,codex,kimi,glm CLIs), each spatially organized in the edges or one over the others.
Then I loop through the desktops checking all the sessions on all the different projects.
This is my active work. Additionally on side projects I got systems and agents working 24/7 that don't require my attention.
This is the important part. Adding this simple rule in your global rules file. For every agent and every repo will improve your focus and decision making by 10x.
Here's the rule, stripped of anything specific to my setup so it drops straight into your CLAUDE.md / AGENTS.md / .cursorrules / system prompt:
## Turn-end brief
End every turn that needs something from me with exactly this block — four
bold labels, each on its own line, its content on the next line, and a blank
line between blocks:
**▍I NEED**
<one line, ≤90 chars: the decision, approval, or input you need — or `nothing`>
**▍I DID**
<one line: what changed · tests/types/lint state · review verdict>
**▍OPTIONS**
1. <recommended option> ← recommended: <≤8 words why>
2. <option>
3. <option>
**▍IF YOU DON'T ANSWER**
<what you will do if I don't reply>
Rules:
- One idea per line. Never let a line wrap (≤90 chars). No horizontal rules.
- Options one per line, recommended first, with the reason it's recommended.
- "IF YOU DON'T ANSWER" must be honest. For anything irreversible — merge,
deploy, publish, money, auth, deleting data, changing a contract someone
else consumes — it is always "I wait." Never proceed on silence there.
For independent work already in scope, "I continue with <next task>" is fine.
- Turns that need nothing from me end normally, with no block. Don't spam it.
same base model. 50% better on their coding bench. thats the whole announcement
GLM 5.3 didnt get a new brain.
They scaled post-training for a month on the pipeline they shipped with 5.2. long context. rl on long-horizon work. async training
zhipu just stood on 5.2 and scaled post-training until coding and long-horizon work moved a lot
-50% on their own code bench
-terminal bench 3 from almost nothing -> 28.3
-deepswe 46.2 -> 66.9
-cyber stuff they didnt even aim for that hard and it showed up anyway
i run agents all day so this is the part that actually hit me
they started training on jobs that eat a real engineer for days
full env / docs / clusters
then they measure the speedup when the work is actually done
weights drop in two weeks
Introducing GLM-5.3: Built to Code. Ready for Cyber Defense.
- Top-tier coding and agentic capabilities, achieved through post-training on the 743B base model
- A major leap in cybersecurity, setting a new standard among open models
Tech Blog: z.ai/blog/glm-5.3
1
209
Seeing this eclipse at Luxor/Valley of the Kings is the closest we’ll ever get to what Ancient Egyptians must have felt when they called the sun a god.
Maybe I go back for another trip just around aug 2 next year. Last time didn’t get to see the new cairo museum finished
198
Had to disconnect today to admire this beauty.
Drove +1000km and +12h total
What an astonishing event. Totally and completely worth it. The 1:39min totality felt literally like out of this world.
Glad I followed @QuantumFracture ‘s advice and went for the long drive to get the whole experience and not just stay in the 95-99% areas
📍Trijuece, Spain
97
You know Hetzner for cheap servers
But they just dropped free api access to some top models
Deepseek v4 flash 304b total / 13b active 512k context text only
Glm 5.2 744b / 40b active also 512k
Kimi k2.7 code 1t total / 32b active text + image 262k context
Qwen 35b total / 3b active same 262k / also does images
OpenClaw instances too if you wanna easily run that
Its the new Hetzner Experiments platform free while the test runs. No sla / no backups / not for prod apis.
Can change or vanish overnight.
For spinning up agent pipelines and burning tokens on tests tho this is free capacity on tap until they kill it
@Hetzner_Online
Lowkey the best CLI out there
Waiting for grok 4.6 to launch next week
Grok Build V1.0 is now released. Try it out!
X.ai/build
1
1
355
Wait for the iPhone fold
Set up tailscale into ur vps’s, install termius, buy the logi mobile fold & a foldable keyboard.
You just got the whole setup to work from absolutely anywhere in the world fit in your pocket.
I’m guessing @levelsio will have this setup as soon as it launches





























