inspire others, seek the truth, be the future
Bali, Indonesia
Joined December 2024
- Tweets170
- Following335
- Followers18
- Likes1.6K
A day at the Park retweeted
A good presentation gives people knowledge.
A better one leaves people inspired.
The strongest messages do not end with the last slide. They leave people debating, challenging, and carrying an idea forward.
Read: yesbutbetter.com/dont-just-p…
A day at the Park retweeted
A billionaire literally explained how the entire economy works in 42 minutes better than most $200,000 MBA programs ever cloud.
A day at the Park retweeted
Le 1er ministre de Slovénie vient de DÉTRUIRE l’UE dans un tweet mémorable ! ⤵️
Le coup d’État de Tusk en Pologne est en train de réveiller toutes les forces patriotiques et de se retourner contre l’UE !
Ko smo se odločali za vstop v EU, si ni nihče predstavljal, da bo v kaki njeni članici in uniji nasploh mogoče brez posledic:
1. Razveljaviti predsedniške volitve (Romunija);
2. Ukiniti informativni program na nacionalni RTV (Madžarska);
3. Poslati policijo na Ustavno sodišče (Poljska);
4. Tri tedne pred volitvami poslati v zapor predsednika stranke, ki vodi v anketah (Slovenija);
5. Zasesti nacionalno RTV s policijo in zamenjati vodstvo (Poljska);
6. Prekiniti mandate vodstvu nacionalne RTV in izvesti čistko zaposlenih (Slovenija);
7. Poslati v zapor člane prejšnje vlade brez možnosti obrambe (Madžarska);
8. Podeliti skoraj milijon državljanstev ilegalnim migrantom po hitrem postopku (Španija);
9. Zapreti novo Nuklearno elektrarno in zgraditi plinovod do Rusije (Nemčija);
10. Obdavčiti ogljične energetske vire ob hkratnem zapiranju NE ter preseliti industrijo v države z nizkimi ekološkimi standardi pod pretvezo, da tako rešujemo planet.
Nastala je EU, ki se mora reševati sama pred sabo.
A day at the Park retweeted
My doctor told me my LDL of 110 is not normal. The average LDL in the United States is 111.
He wanted it under 70. He recommended a statin. I asked him one question. Why does my body make 80% of its own cholesterol if it is trying to kill me?
He did not have an answer. But the data does.
A study of 12.8 million adults found the lowest death rate at a total cholesterol of 210 to 249. The guidelines say get below 200. The Framingham Heart Study found that people whose cholesterol dropped over 14 years had an 11% higher death rate. In the Sardinia Blue Zone, people over 90 with higher LDL lived longer than those with lower LDL.
The guidelines are pushing people into the danger zone on the left side of this chart.
A day at the Park retweeted
🚨 BREAKING: BOJ INSIDER YUTO JUST SAID SOMETHING ABSOLUTELY TERRIFYING ABOUT EUROPE.
“The decline of Europe is not a failure of the plan. It is the plan.”
Let that sink in.
The deindustrialization, the debt spiral, the destruction of purchasing power, the relentless capital flight. What if these aren’t isolated policy failures, but the prerequisites for an entirely new financial order?
History teaches us that the greatest monetary transformations rarely happen during prosperity. They happen when populations become desperate enough to accept what they would otherwise reject.
The most frightening possibility isn’t that Europe’s leaders have lost control. It’s that someone, somewhere, believes everything is proceeding exactly as intended.
And Europe may only be the beginning. 👁️
A day at the Park retweeted
I used to think cholesterol was simple. High is bad. Low is good. Take a pill. Problem solved.
Then I read the Korean study. 14.9 million people. Nine years of follow up. No one on statins.
And the result was a U-shape.
Very low LDL. More cardiovascular death. Very high LDL. More cardiovascular death. The safest zone sat right in the middle.
How can both extremes kill you if the number itself is the problem?
Because the number was never the problem.
Low LDL weakens your blood vessel walls. Cholesterol builds every cell membrane in your body. It makes your hormones. Drive it to the floor and the body cannot maintain itself. The vessels get fragile. They rupture.
High LDL in someone with insulin
resistance, chronic inflammation, and a damaged arterial wall is dangerous. Not because LDL attacked the wall. Because it got trapped inside a wall that was already broken.
But here is what the system never tells you.
A lean, active person eating real food can have LDL of 300 and be one of the healthiest people in the room. High HDL. Low triglycerides. No inflammation. Clean insulin. The LDL is high because their body is running on fat and transporting energy efficiently. Not disease. Metabolism.
Same number. Completely different biology.
Two people sitting in the same doctor's office. Same LDL on the lab report. One needs intervention. One does not. But both get handed the same prescription.
Because the system was built to treat the number. Not understand the environment the number lives in.
Stop chasing numbers. Start understanding your body.
A Giant Korean Study of 15 Million People Just Blew a Hole in the “Lower LDL Is Always Better” Story
Imagine tracking almost 15 million adults for nearly nine years and counting every death from heart disease and stroke. That’s exactly what Korean researchers did with 14.9 million people who weren’t even taking cholesterol drugs. What they found doesn’t match the simple message that “the lower your LDL, the safer you are.” Instead, the risk formed a U-shape. People with very low LDL died more often from cardiovascular causes, people with very high LDL also died more often, and the sweet spot sat right in the middle.
For overall cardiovascular death the lowest risk landed between about 90 and 149 mg/dL. Below that range or well above it, deaths went up. The pattern changed depending on the exact problem. For heart attacks and blocked arteries the best zone was lower (around 70–114 mg/dL). For some types of brain bleeds the risk kept falling only when LDL stayed higher. In plain terms, driving LDL as low as possible protected against one kind of death while raising the risk of others.
When the scientists looked only at the higher range (100 to 300 mg/dL) they still saw a clear rise in heart-attack deaths as LDL climbed. Each jump of about 40 mg/dL raised the chance of dying from ischemic heart disease by roughly 26 percent. But that rise was weaker in women and in older adults than in younger men. Even people in their late 70s and early 80s still showed some extra risk from higher LDL for heart attacks and ischemic strokes, yet the overall picture never became “lower is always safer.”
The study is huge, it followed real people instead of just guessing from lab tests, and it only looked at folks who were not already on statins. That makes the U-shape harder to wave away. Doctors still treat high LDL when people already have heart disease or several risk factors, because the top end of the curve is real. But the data also show that pushing every healthy adult toward rock-bottom LDL numbers is not free. Too low carried its own price in this massive population.
The takeaway is simple and uncomfortable: LDL cholesterol is not a one-way street. In nearly 15 million Koreans the safest zone for dying of cardiovascular disease was not the lowest number on the lab report. Both extremes cost lives. That single finding from one of the largest studies ever done keeps forcing the same awkward question—how low is actually low enough, and for whom?
A day at the Park retweeted
The new Copilot is Microsoft’s bet that the AI race is moving from models to products.
It doesn’t need to own the best model if Copilot can choose among them and keep the customer inside Microsoft.
The potential secret weapon is Autopilot (what I’ve been calling Muse for Business). It knows your workflows, sits across the apps and data you already use, and keeps working even when you step away.
It’s a preview of the battle to come in enterprise AI. OpenAI and Anthropic are racing from models into products. Can they build their own version of this? And can they match Microsoft on the boring-but-critical stuff: permissions, identity, auditability and control, as agents become more autonomous?
My full conversation with @satyanadella. We also talk open vs. closed models, US-China Summit, regulation and the infrastructure buildout.
00:00 Microsoft’s new Copilot
02:02 Why not just give us Autopilot?
05:18 Who pays for always-on AI?
08:06 Copilot picks the model
10:27 Chinese models and OpenAI’s lead
12:04 Competing with OpenAI and Anthropic
13:17 US–China AI talks
16:11 Does AI need new rules?
18:05 The data center backlash
20:40 Is AI being overbuilt?
23:44 Keeping humans in control
A day at the Park retweeted
Groq Founder @JonathanRoss321 shares the biggest leadership lesson he learned from Nvidia CEO Jensen Huang:
“There is no circumstance where Jensen has one-on-ones with people.”
“When you're leading groups of people, if you want to reduce the amount of politics, stop having one-on-ones. Have big meetings with everyone who you want to tell something to and tell them all at once.”
“Copy everyone on the email.”
“If someone says, ‘Hey this person is screwing up,’ copy that person on the email. Let them jump in. Otherwise you're allowing politics to happen.”
“There's no politics. It's the least political large organization you will ever see.”
A day at the Park retweeted
Morgan Stanley sees data center power demand growing ~9x to nearly 79 GW by 2029.
At ~$25M of annual compute revenue per MW, that implies up to $2T of compute revenue and makes entire power stack a picks and shovels opportunity for the AI buildout:
• Grid equipment: $GEV, $FPS, $HUBB
• Busway & liquid cooling: $VRT, $ETN
• Battery storage: $EOSE, $TSLA, $FLNC
• Cooling & CDUs: $VRT, $TT, $MOD, $JCI, $NVT
• Energized sites: $IREN, $CIFR, $CORZ, $APLD, $WULF
A day at the Park retweeted
NEW: Bambu Lab just launched its first laser cutter.
R1 is built to make cutting and engraving wood, acrylic, leather and other materials feel as easy as using a 3D printer.
-Automatically aligns the laser before every job
-Camera shows exactly where your design will land
-Cuts and engraves wood, acrylic, leather and more
-Auto-feed system handles projects nearly 10 feet long
-Rotary attachment engraves bottles and other round objects
-Bambu Suite handles setup, positioning and cutting
Starts at $2,499.
A day at the Park retweeted
CO2 was never in excess. In fact, life thrived during periods of higher CO2 levels. The only thing that is excess is climate ignorance.
unavailable
A day at the Park retweeted
Introducing Sherpa: the most advanced fiction writing AI
We accelerated from $250M in ARR to $500M because Sherpa helped increase content production by 1200% in 1 year
Sherpa was trained on 5.5B hours of playtime with minute by minute dynamic retention data.
550K+ creators have produced 2.6M hours of content annualised using it
Pocket FM is like Netflix for audio-only dramas, with our own pool of one-person studios.
10% of eligible writers on Pocket FM make >$200K
One blockbuster produced >$100M in revenue
3 writers have become millionaires in <2 yrs
We built Sherpa to enable anyone to make >$1M by writing world-class fiction stories:
1. The Idea: Drop a 1-2 sentence concept. Sherpa interrogates it like a veteran editor on tension, stakes, and psychology
2. World & Characters: It builds out the complete lore, tone, and character psychologies
3. Sub-Plot planning: Breaks the premise into arcs, arcs into episodes, and episodes into scenes
4. Scene-by-Scene Generation: Outlines and drafts entire episodes, with you able to steer, rewrite, or override anytime
5. Editorial Review: Stress-tests every draft for pacing, engagement drop-offs, prose, and coherence before it locks
6. One-Tap Production: Pick a voice, convert to audio drama, and publish directly to Pocket FM’s millions of listeners
7. Global Scale & Monetization: Revenue-share on performance, with automatic localization so you earn across international markets
Test Sherpa for free here: pocketfm.com/sherpa
_____________________________________________
Generic LLMs fail at serialized fiction because they lack a long-horizon narrative reward function.
Sherpa solves this through three core technical leaps:
1. Narrative World Model (State Tracking & Retrieval): Context windows degrade over long runs. Sherpa constructs an evolving semantic knowledge graph tracking character states, secrets, and plot dependencies. High-speed retrieval surfaces exact context on demand, maintaining zero continuity decay across hundreds of episodes
2. Hierarchical Story Planner: When writing a 500-episode story like Naruto, you need to plan 100s of sub plots. Rather than generating linearly, Sherpa decomposes narrative across discrete levels: season -> arc -> sequence -> episode -> scene.
Rather than generating everything upfront, like a generic LLM, Sherpa uses progressive planning and dynamic replanning. As the story evolves, it identifies what changed, traces the downstream impact, and replans only the affected parts.
3. Prose Engine (Trained on series' retention data): LLMs write robotically, but serial fiction needs emotion, tension, pacing, and dialogue that sounds like real people.
Sherpa's Prose Engine was designed specifically for storytelling. It was built on 1B+ tokens of Pocket's own stories, trained by learning from what listeners engage with, where they drop off, and what keeps them hooked.
Feedback is taken from specialized evaluator models that measure every scene against a 40-item checklist. (Evaluator models were benchmarked against human reviewers and matched them 80–90% of the time.)
_______________________________________________
Owning distribution and creation puts us in a very unique spot.
More shows -> More data -> Sherpa becomes better -> more creator success -> more creators -> more shows
Pocket FM has already seen one $100M IP. I believe Sherpa will soon lead to dozens of single-person studios creating billion-dollar shows.
Most people are scared of AI but I think it'll unlock more human creativity, help creators earn more, and bring the next great IPs to life. This will create millions of jobs and new income streams.
A day at the Park retweeted
Jev Founder, Diogo Almeida (ex-OpenAI):
"The next era is not the Claude Code era. Claude Code is still part of the assistance era - Jev is the next step
Jev is x200 faster, x400 cheaper and has 0 hallucination. This is what comes after RLHF"
In 17 minutes he explains why every model you use has a human-shaped flaw baked into it and why Jev is better
worth more than a $500 Claude Code engineering course
watch today, then read how to actually boost yourself x100 with Jev in the post below
How to use Jev, and where it actually gives you the 100x:
setup takes 10 minutes:
1. join the waitlist, people are getting approved same day
🔗 typesafe.ai
2. install the official skill so your agent writes correct calls:
- npx skills add typesafe-ai/skills --skill typesafe-ai
on Claude Code it's two commands, the marketplace add on its own doesn't install anything:
- claude plugin marketplace add typesafe-ai/skills
- claude plugin install typesafe@typesafe-ai
3. create an API key in the dashboard
4. in your prompt just say: "use the TypeSafe skill"
now the part nobody is posting:
the 100x isn't the model, it's where you put it
you don't get it by swapping your LLM for Jev
you get it by deleting the calls that never needed a language model
open your agent and find every call that just picks something:
> which tool next
> is this spam
> is this chunk relevant
> does this need a human
> is this diff risky
none of those are writing tasks
they're if statements you outsourced to a frontier model
here's the upgrade, in order:
1. replace each one with a typed question
Choice picks from up to 255 options, Score places it on a 2-10 level scale, Noul returns a raw 0-1
2. batch them
questions in one call run in parallel and barely move the latency, and output tokens are free
so ask every question you might need, including the ones you'll throw away
3. threshold on confidence, not on the answer
under 0.5 escalate to a big model or a human
0.85+ before anything irreversible
4. never let it invent options
build the candidate list in code, from the DOM, the retriever, the tool trace
then let it pick
5. put it in the loop, not next to it
router picks the cheap model, gate checks the tool call before it runs, judge verifies the output after
that's where the heaviest calls in your agent are hiding
6. start with compaction tonight
score every tool call, drop the dead ones, keep the survivors verbatim instead of a lossy summary
lowest effort win available and you'll see it on tomorrow's bill
the honest part:
text only right now, no images, no audio
and on broad benchmarks it loses to frontier models
but somebody ran 18,514 emails through it zero-shot and got 98.33%
against a TF-IDF classifier trained on 14,800 labelled examples that got 98.39%
no training data, $1.12 total
it wins on narrow, well specified decisions
which is most of what your agent is actually doing all day
today gonna share use case how i integrated it to content creation and how i find winning meta ads now in a seconds...
A day at the Park retweeted
THIS CHINESE TRADER USED CLAUDE TO BUILD AN AI TRADING BOT
And recorded this tutorial where he shows how to replicate it from scratch
23 minutes. Step by Step & Free.
Save this before you forget it
A day at the Park retweeted
I think we'll see 5x the number of people become hardware and robotics founders over the next 18 months
Why? Because you can rent every step, from design to manufacturing:
DESIGN
- Astra to drive Blender, FreeCAD, and KiCad the way a person would, so you get editable geometry instead of a dead render. Zoo dot dev's text to CAD API if you want it programmatic.
- Or have Claude write build123d code, which is Python parametric CAD, and there's a Claude Code plugin that runs it and exports the STL for you.
PROTOTYPE
- Bambu A1 mini is $299, P1S around $399 on sale. Or skip owning one, JLC3DP prints and mails it for about $20. PCBWay does resin, SLS, and CNC if plastic won't cut it.
ELECTRONICS
- KiCad for the board, Astra can drive it. JLCPCB fabricates and assembles, often under $100 for a small run. ESP32 for wifi, Raspberry Pi if it needs a brain.
ANYTHING THAT MOVES
- Unitree sells a Go2 with a full SDK for about $2,500.
- Hugging Face put out a $399 open-source biped.
- LeRobot gives you the whole train in sim, deploy to real pipeline, and Physical Intelligence open-sourced π0 so you're fine-tuning instead of starting from zero.
- Prototype the policy in MuJoCo or Isaac Lab first.
MANUFACTURING
- Alibaba RFQ for the first 100 units. Check 1688 to see what the factory actually charges domestically, then negotiate. Pietra or Sourcify if you want someone to handle it.
FULFILLMENT AND SELLING
- ShipBob or Amazon FBA. Shopify for the store, TikTok Shop for distribution, Kickstarter if you want the money before you build it.
How to think about starting your own robotics or hardware company:
1. Pick a niche that's already buying weird gear. Cyclists, tabletop gamers, beekeepers, home baristas, dog people with mobility issues. These groups spend money on specific objects and complain in public about what doesn't exist. Or grab ideas off Ideabrowser.com
2. Go read the complaints. Reddit, IG etc Search "I wish someone made," "does anyone make," and "modified my." That last one is the best signal, because someone already hacked the product together and you're just manufacturing what they built by hand.
Also check Etsy!! If 3 sellers are doing a janky 3D printed version with 400 reviews each, the market is validated.
3. Make one. Describe it to Astra or Claude, get the CAD, print it in ugly gray PLA, use it, fix it.
4. Only go to Alibaba or similar once you've sold a few. Message 10 suppliers through RFQ, take the third cheapest, always pay the $50 for a sample before the real order.
5. Film everything from day 1. The first ugly print, the failed version, the box of 100 arriving. That's your entire marketing budget, and hardware is one of the few categories where people actually want to watch the thing get made.
6. Raise the price. Almost everyone here anchors on what the plastic cost. Your customer is comparing you to nothing, because the alternative is the product doesn't exist. Start at 5x COGS and go up.
THIS GOAL OF THIS POST IS JUST HERE TO GET YOUR CREATIVE JUICES FLOWING. Of course, you can build robotics/hardware in a bunch of different ways.
One thing I've learned is data couldn't be more important when you're building a hardware/robotics startups.
These models learn from first person video of a human doing the task, and that footage doesn't exist for almost any job.
So basically you pick one repetitive job people quit over, film someone doing it for 2 weeks, fine-tune π0 on that footage, use organic to sell the first few units and figure out scaling,
5 years ago you needed a factory, a supply chain, and a $1M just to find out if anyone wanted the thing.
Now, anyone can become a hardware/robitics founder.
And I suspect a lot of people will become one!
Vibe manufacturers.
