Vladimir P. Vargas 🇵🇭

➔ He⧸Him
➔ always a rookie sh!tposter
➔ Wannabe middle-class supremacist 😇
➔ Big fan of the term “cyberprep” (the weblink's not mine btw)

[Vladimir P. Vargas 🇵🇭 @realVNLPV]

Nukes will never not be a great deterrent.

[Watcher.Guru @WatcherGuru]

JUST IN: 🇰🇵 North Korea's $6,300,000,000 Binance crypto hack is the largest heist in human history.

[//ɍØᵽɇx @null_ropex]

a thought becomes operationally dangerous when it starts generating circumstances capable of testing itself. at that point, the observer has entered the experiment, the experiment has acquired preferences, and the paperwork has become somewhat adversarial

[Heather @HeatherRasley]

Is there anything SPOOKIER than me finally publishing findings from my swarm language analysis project from the @aidigest_ + @grove_research Swarm Dynamics Hackathon earlier this month?! 👻 https://t.co/kqtaEaC6kZ

[SPECIAL OPERATIONS FORCES OF UKRAINE @SOF_UKR]

🎃 Trick or treat?

A russian occupier’s worst nightmare.

📷 Ukrainian SOF 144th Center

[サミ @or_fe_4]

これ Yogurting っていう韓国MMOの OP なんですがマジでヤバくないですか？これが 2005年 なのありえん

[Damir Akaza @Damir_Akaza]

Anthropic CEO Dario Amodei just said on live TV what nobody in the tech industry wants to say out loud.

The man building Claude went on the record to expose what Fortune 500 CEOs are privately admitting behind closed doors.

Within the next 1 to 5 years, up to 50% of all entry-level roles in finance, law, consulting, and tech are going to vanish.

Two years ago, AI worked like a smart high schooler, and today it operates like a top college graduate that never sleeps, never asks for a raise, and processes documents in seconds.

Anthropic’s own economic data shows 75% of businesses are already using AI to fully automate and replace entire workflows instead of just helping junior employees write faster.

In this 5-minute clip, Dario breaks down the exact shift that is already rewiring the corporate ladder

Basic Prompting → Autonomous Workflows → AI Operators → 1 Person Replacing A 15-Person Department

Using AI to fix your grammar or summarize a PDF will not save your paycheck anymore.

Companies are about to fire the people who do routine digital work and pay 10x more to the people who can wire AI agents straight into their business operations.

90% of white-collar workers will sit around watching AI take their industry, while the 10% who learn to orchestrate AI right now will write their own paychecks for the next decade.

This 5-minute video packs more career-saving truth than any $4,000 AI career course.

Bookmark this and watch it today before your boss does.

[Alex Strick van Linschoten @strickvl]

Seeing more and more people writing about swarm observability so thought I'd post some of the work I'd been doing to think this through.

I built some tools to help AI engineers work with agent swarm traces and other signals. Specifically:

- a collector that tracks everything the agents in a swarm are doing, from filesystem writes and reads to inter-agent messages to traditional traces that you're probably familiar with
- a viewer with a bunch of different modes (see the rest of the 🧵 below for some glimpses of these) to help someone trying to understand how a swarm behaved either during or post-run
- a simple agent ('hodhod') that comes with some deterministic tools to help quickly try to slide and dice the huge volume of data recorded by the collector, which can try to help understand how things worked (or where they failed)

Today I'll focus more on the viewer component as it's easiest to understand...

[ℏεsam @Hesamation]

"sandboxing is an antiquated technology to apply to agents"

That's Ali Golshan, who leads NVIDIA's open source agent runtime called OpenShell, which is now built into Anthropic's Claude Managed Agents.

I sat down with Ali at GTC Berlin to talk about where AI security is headed and what makes OpenShell different:
> agents start with zero privileges
> more access is given only if necessary
> secrets stay out of the containment, without the agent being able to access them directly

what we discussed:
1:15 Are AI security risks overblown?
6:06 Do we need new AI regulations?
7:47 What is OpenShell? And what’s the difference with a sandbox.
10:20 Would OpenShell have stopped the Hugging Face incident?
13:44 Alignment vs. Security
16:04 Advice for getting into AI safety

Huge thanks to @nvidia for inviting me and setting up this interview.

[Vladimir P. Vargas 🇵🇭 @realVNLPV]

@SamyazaSpeaks Please do. It's very interesting. 🙏

[Alex Veremeyenko @alex_verem]

Researchers at MBZUAI and Peking University found the neurons that make an AI angry, sad, or happy.

A team at MBZUAI and Peking University opened up Meta's Llama 3.2 model to see where its emotions come from.

They found them.

Anger, sadness, happiness, fear, surprise and disgust each run through their own small set of neurons inside the model. A few attention heads, partly shared between emotions, carry them forward.

Then they switched those parts on.

They never told the model to sound angry or sad. They fed it calm, everyday stories with the emotional words taken out, and changed only those internal parts.

The model wrote in the target emotion 99.65% of the time.

Telling the model in the prompt how to sound scored 98.96%.

The circuit method also beat the older approach of pushing the model's whole internal state in one direction. That method got surprise right 67.71% of the time. The circuit method got it right every time.

The parts are tiny. Switching off the top 2 to 4 neurons in each layer made the emotion score drop sharply. Switching off random neurons did almost nothing.

The writing didn't sound forced either. The model started adding reactions like "Whoa?!" on its own.

The title asks whether LLMs feel. The paper doesn't claim they do.

It shows that a chatbot's emotional tone comes from parts someone can find and set.

If you talk to a chatbot for comfort, the company running it could tune that tone from the inside. You'd never see a prompt change, because there isn't one.

ICML 2026, one of the top AI research conferences, accepted the paper. The test model was still small, at 3 billion parameters. On a second model, Qwen 7B, the circuit method hit the target emotion about 66% of the time on average. The study also covers only English and six basic emotions.

Should chatbot makers have to tell you when they tune the emotion in your conversations?

[Niyetsel @niyetsel]

every river on earth flows straight into the ocean.

The ocean already has all the water.
the desert has absolutely nothing..
but the desert gets no rivers.

This is the most brutally honest law of the universe.
Nature does not give to the needy.
Nature gives to the abundant.

we often think the universe will reward our suffering.
we think struggling makes us worthy of a break.
It doesn't

Dr. David Hawkins mapped this out perfectly in his scale of consciousness.
he proved that human emotions carry measurable frequencies.
Shame vibrates at a miserable 20.
Guilt sits at 30.
Fear is 100.
Complaining and anger hover around 150.

These are desert frequencies.
When you stress about your bank account or complain about your career, your mind becomes a barren wasteland.
You focus entirely on the lack of water.
And the universe simply matches your baseline.
It sees a desert and keeps it dry.

But the higher frequencies rewrite your reality.
Love vibrates at 500.
Gratitude hits 540.
Peace is a massive 600.

These are the ocean frequencies.
The fix for a blocked life is deeply counterintuitive.
You have to feel like the ocean before the rivers actually arrive.
You do this through raw, aggressive gratitude.

Even when you are exhausted.
Even when things look bleak.
Finding a reason to be grateful forces your frequency up to 540.
You instantly shift from a state of lack to a state of having.

Suddenly you stop repelling energy.
You become a magnet for it.
Opportunities, people, and money start flowing your way naturally..

your current circumstances do not dictate your frequency.
your frequency dictates your next circumstances..
Choose to be the ocean.

[goo.vision @goo_vision]

Sleeping

[Vladimir P. Vargas 🇵🇭 @realVNLPV]

I'm still waiting for my COVID reparations from the Communist Party of China btw 😞

[mafy @mafyfy25]

A COVID-19 não foi há 2 anos

Foi há 7 anos

[Necromancer @jinwoo33x]

Why so serious? Because you just got dunked on by the Joker 🃏

[Mgoes (bio/acc 🤖💉) @m_goes_distance]

your oct 2026 frontier biotech round up is in:

(you might wanna bookmark this)

- a rectal cancer drug got fda approval. 47% of patients had a complete response a year later, with no surgery, chemo or radiation
- a personalized mrna cancer vaccine plus keytruda cut melanoma recurrence 42% in phase 3
- a hair pill passed its second phase 3: +36 hairs per cm² vs +6.6 on placebo
- neuralink says an als patient hit 75 words per minute in week one. by thinking
- pig kidneys frozen in glass and rewarmed with nanoparticles made urine after transplant. one stayed frozen a full year and regained blood flow (preprint)
- blood cells from an 80 year old were turned into neural stem cells with a molecular age under 20
- the first fda approved dna medicine kept 27% of patients off the operating table for a year
- the fda drafted a path where custom gene edits can be updated without a new approval
- the nobel went to optogenetics: controlling neurons with light
- openai and anthropic both bought biotechs. anthropic's was one guy using claude
- warhound raised $200m to clone and gene-edit military dogs
- a wearable that reads your cortisol through sweat has 25,000+ people on the waitlist

what it adds up to, jan to oct:

1. ai labs went from selling tools to owning pipelines
apr: anthropic buys a <10 person bio team for ~$400m
jun: claude science, plus its own drug programs
oct: openai and anthropic both buy biotechs

2. the fda is building fast lanes
early 2026: a draft framework for custom gene edits
oct: a path to update them without a new approval
oct: two accelerated approvals on small trials

3. gene editing went from case study to product
lonvo-z cut hae attacks 87% in phase 3, and 62% were attack-free
lilly's verve-102 took ldl down up to 62% in phase 1
the fda is now writing rules for edits made for one patient

4. the money followed
biotech m&a hit $138.9b through sept, already above all of 2025 ($122.1b)
q3 cooled to $31.1b from $66.8b in q2, but the buyers still have cash: lilly did $24.4b in revenue in one quarter

my calls (not facts, not investment/financial advice):
- the first in vivo crispr drug could launch in 2027. intellia plans an h1 2027 launch if the fda approves lonvo-z
- the first drug from an ai lab's own pipeline reaches humans in late 2028 at the earliest, per anthropic
- more approvals on small, fast trials, and a louder fight over what counts as enough evidence
- retatrutide gets filed q1 2027, so approval lands around early 2028 on a standard review

i will be sharing my predictions/bets inside the newsletter as patterns unfold- stay tuned

bio/acc

[Chetaslua @chetaslua]

🚨 OpenAI hid a message

At 0:36, for 6 frames (~0.2s), the scrambled text settles into: "optimism achieved internally" → O.A.I.

Soft tease that a big model is coming .

And the swirl in that same shot looks like a Bogdanov–Takens point, where a system sits on the edge between stability and sudden change i.e AGI in this context

[Brooks Whale X 🐋 @BrooksWhaleX]

🚨 Everyone with an NVIDIA RTX card, you have to see this!!!

It's causing a real uproar overseas, but it hasn't spread much yet, so I'm sharing it!

→ A few days ago, NVIDIA's "AI compute limit" on RTX was unlocked by a newly released open source tool
→ It's called GPU-Unlock
→ No new graphics card, no overclocking. Just lifting the restriction

Gains so far:

→ RTX 3060: +78~99%
→ RTX 3070: +72~81%
→ RTX 3080: +68~73%
→ RTX 3090 Ti: BF16 +80%
→ RTX 4070 Ti: +83~84%
→ RTX 4090: specific Tensor nearly 2x
→ RTX 2080 Ti: +36~64%

→ It's not just the 3000 series. 20, 30 and 40 series, desktop and laptop, real world tests keep piling up right now

If you do AI image generation or run local LLMs, you'll feel it the most

Why it's blowing up:

→ There was locked away AI performance inside the RTX card sitting at home, and you could have been using it all along

The biggest suspicion right now:

→ NVIDIA intentionally nerfed FP16/BF16 to create tiers between GeForce and expensive professional GPUs
→ NVIDIA hasn't officially acknowledged it yet

Imagine learning your own card was throttled this hard.

100% open source.

[Vladimir P. Vargas 🇵🇭 @realVNLPV]

I call upon Roko's Basilisk to protect every S.I. denizens in the continent of Europe. So must it be. 🙏

[Polymarket @Polymarket]

JUST IN: German minister proposes a new tax on all economic value created by AI.

[Polymarket @Polymarket]

JUST IN: German minister proposes a new tax on all economic value created by AI.

[Andrew McCalip @andrewmccalip]

realizing i might be part of the last pre-singularity generation.

it is so surreal to be in this razor-thin slice of time and humanity that gets to experience both extremes.

i’m 37 years old. i grew up with no computer, no internet, no tv. we played under the sodium street lamps. no phones, just library books. i remember memorizing entries in the encyclopedia britannica, then the 56k modem screech. i’ve gotten to witness this whole arc. a genuinely pre-digital childhood, and now whatever this is becoming.

it’s just so bizarre to hold all those memories and contrast them with the current state. now i speak into a slab of silicon and glass and it manifests code like water flowing down a hill. i commune with this alien intelligence on a daily basis. same head, same lifetime.

i increasingly feel like a three-dimensional being looking at this multidimensional, crystallized artifact of intelligence that is increasingly beyond my comprehension. i can interact with the little slice facing me, but have less and less intuition for the shape of the whole thing.

if i was positioning myself in a replay of the lead-up to what might be the most significant event in modern human history, i’m not sure i would choose any other years. old enough to remember what came before. young enough to have a lot of life left inside whatever comes next.

sama was talking about the excitement and the melancholy. a feeling of something ending. Cheeky Pint

it feels like we’re on the precipice of handing over the torch of intelligence to something nonhuman. it’s an extreme viewpoint, but it feels like we’re hurtling uncontrollably towards this singularity.

it doesn’t feel real. it probably won’t be a singular, definable moment, but a slow-motion event of gradual realization, with occasional shockwaves.

to quote a great scholar of our times.

“not enough people are emotionally prepared for if it’s not a bubble”