Lets traditional startups supply equity shares on blockchain.
Joined April 2017
- Tweets2.1K
- Following1.6K
- Followers587
- Likes4.2K
Fundaria retweeted
We just built a neural adapter that uses the real architecture of human neurons from the H01 connectome to construct virtual brain tissue, then wires that virtual tissue directly to living CL1 cultures (real human cortical neurons on a chip).
It drives the actual behavior of MLP neurons inside models like Qwen.
Silicon MLPs + H01 morphologies + living wetware.
The bridge is open, let dive into how it works 🧵👇🧠⚡️
Fundaria retweeted
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
Fundaria retweeted
💫 PROMPT SHARE 💫
If the entire universe were generated by prompts, and you could inspect the original prompt that instantiated me, what would it be?
Fundaria retweeted
想了解臭名昭著的朝鲜黑客组织 Lazarus，可以看看 Cipher 拍的纪录片《How North Korean Hackers Are Stealing Billions》。
Cipher 专做犯罪和地缘政治题材，看他们的纪录片跟追剧一样。这部片长 1 小时 20 分钟，把这支朝鲜国家队从头到尾讲透了。
Lazarus 可不是普通黑客，它是朝鲜国家养的「犯罪企业」。大多数黑客都是从小被选中的数学天才，每天工作 15 到 20 小时。Lazarus 的 KPI 只有一个：给政权搞钱，大头流向核武。
片子从朝鲜战争、主体思想和饥荒讲起，先讲清楚这个国家为什么会养出这么一支队伍。
然后按时间线一路往下：
① 1980–2000 年代伪造超级美元；
② 2014 年黑索尼影业，起因是那部《刺杀金正恩》；
③ 2016 年孟加拉国央行 SWIFT 劫案卷走 8100 万美元，还让打印机打不出转账记录，好拖延被发现的时间；
④ 2017 年的 WannaCry，用的是泄露出来的 NSA 漏洞「永恒之蓝」，波及约 150 个国家、20 多万台电脑；
⑤ 转战加密：Coincheck、Ronin (Axie)、WazirX，2025 年 Bybit 被盗 15 亿美元，再到现在 2026 年的币安。
------
币圈钱多、盗完还有机会全身而退，Lazarus 只会一直盯着挖漏洞。行业是时候团结起来了，我们面对的是一整个国家机器。
Fundaria retweeted
Sorry for the wait, the episode is here!
Live action with real robots instead of animation (episode 4): "Anger"
Enjoy!
As always, the software of Microduck and Reachy Mini is open source:
github.com/pollen-robotics/m…
github.com/pollen-robotics/r…
Fundaria retweeted
Physicists in Okinawa have floated a diamond about a centimetre wide and pushed it up and down using nothing but the spins of electrons inside it, pumped by pulses of green laser light.
That object is roughly a hundred million times heavier than anything moved by spin force before. So, a quantum force now shows up directly on something heavy enough for gravity to matter.
Fundaria retweeted
Holy fuck... Slopcore animated manga @anabology . All done with Agents running in the background doing their thing.
Fundaria retweeted
How does bamboo turn into clothes
Readers added context they thought people might want to know
The video is AI-generated slop that inaccurately depicts bamboo-to-clothes production; most commercial bamboo fabric is chemically processed viscose/rayon from pulp, not direct mechanical fiber.
scienceinsights.org/how-to-make-ba…
ftc.gov/bamboo-textiles
x.com/LEARNDAlLY/sta…
Fundaria retweeted
👇 BEST EXPLANATION EVER ON HOW LLMS ARE BUILT
it's surprisingly simple when you break it down!
1️⃣ Gather and clean enormous amounts of data.
2️⃣ Convert text into tokens and vectors.
3️⃣ Train transformers to recognize patterns.
4️⃣ Predict the next token billions of times, improving each time.
5️⃣ Refine with feedback and reward signals.
6️⃣ Compress and optimize for real-world use.
but here's the fascinating thing:
nobody hardcoded the ability to write software, solve problems or explain science.
much of it emerged from one simple task: predicting what comes next.
and we're still discovering what else they can do.
Fundaria retweeted
I just made a 908-page comic book.
19 chapters, dozens of consistent characters, all adapted from a 95,000-word novel.
Here's "The Hobbit" by J.R.R. Tolkien, turned into images.
How I built it: reference images, agents and a tight harness 👇
Fundaria retweeted
1/ We discovered something totally unexpected about the neural structure of consciousness that overturns decades of thinking on its biological basis and brings us significantly closer to the recipe for generating conscious awareness.
New preprint: biorxiv.org/content/10.64898… (led by @HesseJanis, with @FrankLanfranchi, @_conor_mcgrory, @Yuelin_Shi, and @EngelTatiana)
