Daniel Matos retweeted
Introducing Analisa USA 🇺🇸
One free map connecting how Americans live, where federal money goes, how Congress votes, and who is running in 2026.
Every county, contract, roll call and race, linked to its official source.
tryanalisa.com 🧵
Daniel Matos retweeted
For whatever it's worth, there is no other platform in the world where the CEO of the third largest company in the world personally responds to a request from his biggest competitor about listing on their platform.
Competition aside, let's take a moment to appreciate what a special platform X is.
I keep telling my friends this. X has the highest % of influential people actually running their own accounts. Trust me, this is actually Sundar typing on his Pixel, replying to Peter. What an amazing place.
Daniel Matos retweeted
Replying to @levelsio
Like most of the high end mobile phones at the time, the Nokia 9110 had an Infrared port for close range communications.
This was ~1998. Bluetooth would come only a couple of years later.
But the 9110 allowed you to run custom apps and the IR port was accessible as a serial port.
Around the same time, Citroen had cars with an IR key fob. It had a fixed code so it could be easily cloned.
So, as far as I know, the humble Citroen Saxo was the first car that you could open from your mobile phone 😃
Daniel Matos retweeted
Opus 5.5 designing LEGO 👀
I asked it to design a Microduck I can build with real LEGO pieces. It:
> designed it life-size using 1113 real LEGO parts
> verified: 3,204 connections, 0 collisions, every step buildable, centre of mass inside the feet 🤯
> made a 141-page LEGO-style booklet (237 steps)
> priced every piece in the browser and prepared the orders on BrickLink
Estas apreensões diárias de jerricans de combustível fazem-me lembrar a anedota do gajo que passava a fronteira todos os dias com uma mochila carregada de areia.
Daniel Matos retweeted
12M+ views so I had to dig into this
Is it just a hyped launch or something novel and actually useful?
I’m leaning to the latter and will break it down with a practical example
First, understand that Jev isn’t a typical LLM that you prompt and get back a written response
It’s much more specific - you give Jev context along with a set of questions and possible answers and it returns probabilities to each question and answer set, and it does it super fast and cheap
I found it helpful to think through a simple example:
Let’s say you’re an online retailer and you get hundreds of inquiries a day
You put an agent in the workflow that looks at each inquiry and makes a judgement call on the next action and then generates a response
But that LLM judgement call can be messy - LLMs still have a tendency to make shit up or respond in ways that you don’t like, so we put humans in the loop
What Jev does is look at each inquiry and makes a fast decision on the nature of it so the workflow can take the next best action with confidence
It doesn’t replace the workflow or the LLM entirely, it handles the fuzzy fork in the road that used to require a human to read the message first
Let’s say you get a customer email:
“This jacket sucks! The zipper jammed the first time I wore it and now it won’t close so I want my money back.”
That message lands in the retailer’s system along with a few facts the company already has: the order is 11 days old, the return window is 30 days, the jacket was $100, and this customer has one previous order with no refunds
In a typical LLM workflow you’d let the model decide what to do and have it initiate that action and send a reply email, BUT many companies want a HITL in many of these cases, especially ones where it’s not entirely clear what to do
Imagine the LLM replies to the email with troubleshooting instructions on how to fix the zipper - probably not the best action
With Jev, the company doesn’t ask it to write a reply, it asks a few specific questions about the email and order
>Does the customer want a refund?
>What is the main issue: defective product, didn’t like it, late delivery, or something else?
>How unhappy does the person sound?
>Does this look like it falls inside the published return policy?
>Should this go to returns, quality control, or a human agent?
Jev looks at the email and the order context, then answers all of those at the same time with a probability score - no prose, just a score
In this case the picture comes back fairly clear: yes, this is a refund request, the issue is a defective product, frustration is high but not explosive, the request is inside the return window, returns should own it
That is enough for the retailer’s existing rules to take the next action with confidence
If Jev had been less sure - say the email was sarcastic, the order was 45 days old, or it was unclear whether the customer wanted a refund or a replacement - the workflow could put the ticket in a human queue instead of guessing
The useful part is the handoff, specifically knowing when to pass it to the next step in a workflow or route to HITL
If Claude were the judge in a less clear situation, you’re likely to get a long winded answer that still leaves you wondering what to do next, and it would have taken longer and cost a lot more
This was a simple example but I can see how this can be incredibly useful in complex enterprise workflows which is why I lean towards novel and useful over just a hyped new thing
Will be following this one closely
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
Daniel Matos retweeted
this is the easiest way to understand Jev:
LLMs generate answers.
Jev makes decisions.
that sounds like a small difference, but it actually changes the entire use case.
say you give a normal LLM this:
“here’s a user, their account history, payment behavior, support chats, device data, etc.
tell me if this looks risky.”
the LLM might reason through it and return:
“yes, this looks high risk.”
maybe in JSON if you ask nicely.
with Jev, you define the possible decisions upfront:
risk:
* low
* medium
* high
manual review:
* yes
* no
and Jev returns something closer to:
risk = high (96%)
manual review = yes (91%)
that’s basically the product.
it’s not trying to be another ChatGPT.
it’s more like an AI-native if statement.
instead of:
if transaction > $10,000:
review()
you can start thinking more like:
if “does this behavior look suspicious?” > 95%:
review()
and that opens up a pretty interesting category of software.
a few assumptions I had at first that turned out to be wrong:
1. “so it’s just a classifier?”
kind of, but that undersells it.
the input can be messy real-world context, and you can ask multiple typed questions about that state at once.
fraud?
churn?
escalate?
eligible?
priority?
all from the same input.
2. “so it replaces GPT / Claude?”
not really.
I actually think the interesting architecture is:
Jev decides WHAT needs to happen
Claude / GPT reason or generate WHEN deeper intelligence is needed
normal code executes the deterministic stuff.
Jev becomes the routing layer.
3. “it can’t hallucinate?”
this one needs nuance.
if your allowed answers are:
LOW
MEDIUM
HIGH
Jev won’t suddenly invent:
“EXTREMELY HIGH 🚨”
the output structure is constrained.
but it can still be wrong.
HIGH at 92% can still be the wrong decision.
so “no hallucinations” doesn’t mean “always correct.”
4. “why not just force an LLM to return JSON?”
you can.
we already do this everywhere.
but you still deal with generation latency, schema validation, retries, weird outputs, confidence estimation and a lot of glue code.
Jev is designed around the decision itself rather than text generation.
5. “why should I care?”
because most software is ultimately a giant tree of:
if this → do that
if this → route here
if this → escalate
if this → reject
if this → ask a human
Jev is basically asking:
what if those if statements could understand messy human context?
that’s a much more interesting framing than “another AI model.”
I can see this being very useful for:
fraud / risk
support routing
moderation
PR / QA automation
lead scoring
compliance
workflow orchestration
agent routing
especially as the cheap + fast decision layer sitting in front of larger reasoning models.
early tech, obviously.
but the category itself makes a lot of sense.
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
Daniel Matos retweeted
hype-free explanation of jev:
jev does not replace gpt / claude
jev is just a *really* smart switch statement
like if 2016 ml classifiers got 2026 levels of intelligence
it's a new* type of tool that will make a lot of workloads insanely fast, cheap, and accurate
* = and by new, i mean rebranded
~~~
it needs a predefined set of options and it will tell you which one to take
it cannot:
- write code
- generate natural language
- reason step by step / show its work
- produce any output you didn't define in advance
- pick from more than ~255 options in one shot
but it can:
- classify, route, score, rank
- give confidence
- pick the right branch, tool, model, or sub-agent
- judge / verify / guardrail an llm's output
- label tons and tons of rows
~~~
i'd imagine a lot of workflows that look like:
llm proposes options → jev decides → code executes
and i see this fitting *really* well with code mode and mcp
~~~
implying this will lead to agi seems incredibly far fetched to me, but i don't want to discount the types of applications that this will make possible
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?
I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev
• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions
AFAICT the shortest path to AI-based economic revolution
Daniel Matos retweeted
This is the story of FCKGW-RHQQ2-YXRKT-8TG6W-2B7Q8.
I was the primary author of Windows Product Activation on the operating systems side. Not the key math – those were the ACTUALLY smart guys with heads so big that they bumped the doorjamb on the way into your office. I just did the protection of and integration into the operating system.
As you likely recall, Product Activation required that you have a key, and you’d have to activate it online or over the phone.
But Microsoft had large corporate customers, think hundreds of thousands of PCs, and those customers did not want to have to activate every offline PC. So in that case, you would be given special “Volume Media” that could be unlocked with a Volume key.
That way, a volume key could never unlock your retail CD.
Now, to know what media you have, I inspect the disc image. And the Volume licensing media contained a special binary blob that I made. It was about 10MB of data, entropy stripped, hard to duplicate on Usenet at the time. And it was only present on the Volume media.
That secret data, which I hashed and checked to match to the key, was Microsoft Bob. I took their disc data and compressed and encrypted it several times. Rest assured it’s encrypted and even in the AI age no one has extracted it yet. I really just needed random data, so I started with a precompressed binary image because, secretly, I was young and didn’t know if I could trust CryptGenRand
And so, if you ever tried the FCKGW key on a regular XP CD, it wouldn’t work, because you didn’t have Bob buried in your bits.
But if you had the means in 1995 to download large ISO binaries or copy CDs, all it took was the key and a copy of the disc.
The key itself still had to pass the product-key validation math (the part I didn’t own). Once it did, WPA simply skipped the rest of the process. That combination—valid VLK + matching volume media—is what made the install look fully legitimate, including early Windows Update checks.
How it leaked
RTM went to manufacturing on 24 August 2001. Retail launch was 25 October. Roughly five weeks before launch the warez group Devils0wn (often written “Devil’s Own”) posted a complete, non-beta “Windows XP Pro Corporate” ISO together with the FCKGW key. That ISO was the volume media. Pirates then started baking the key into images or just writing it on the CD with a Sharpie. The famous photo of the marked-up disc is from that scene.
The most plausible source is an OEM or hardware partner that received final media early so they could ship PCs on launch day. Dell is the name that comes up most often; Intel has also been mentioned because of later source dumps. There were very few organizations that actually possessed a VLK that early, which is why the timing points at a partner rather than a random corporate customer. But I do not know. Presumably someone does.
Aftermath
We started blacklisting the Product IDs that belonged to the leaked keys (FCKGW decoded to one of the 640-range PIDs we later treated as toxic) in SP1. SP2 and Windows Genuine Advantage made the check more aggressive and also started blocking updates for those installs. Other, un-leaked VLKs continued to work on volume media for years; Microsoft did not kill the entire volume-licensing path, just the ones that had escaped.
So the design decision was “give trusted big customers a convenient offline path.” The operational failure was that the media and one of those keys left the building before the product even reached stores. Once both pieces were public, WPA’s hardware-binding never got a chance to run.
These Days
If you’re curious about what I’m doing today, check out Task Manager TMOG at tmog.org where you can download the Free version! It runs fully native on Windows, Mac, and Linux!
Daniel Matos retweeted
Microsoft has dropped CVE-2026-69730: a Windows DNS Server RCE with a 9.8 CVSS score.
No authentication. No user interaction. Low complexity. No mitigation. No workaround.
Basically, it doesn’t knock on the door; it lets itself in, makes a brew and starts executing code.
Patch it. 😂
Daniel Matos retweeted
En 1986, un ingeniero japonés obsesionado con los coches metió un Ferrari Testarossa descapotable en un videojuego. No era un juego de carreras, ya que no había rivales. Solo una carretera, una chica al lado y música a todo volumen. Se llamaba Out Run. Tira del hilo 🧵👇🏽👇🏽👇🏽
I would go for a pair of these over a Patek Philippe any day!
To celebrate 55 years of the Technics SL-1200 Series, we introduce the Gold Edition SL-1200M7GL/M7G, built on the performance of the SL-1200MK7.
The timeless value of truly engaging with music remains pure gold for generations to come.
#SL1200_55thAnniversary #Technics































