Yells a lot. Just good enough. Rule 1 is always blame me. kissed by lavender.
internet
Joined June 2010
- Tweets17.3K
- Following915
- Followers1.3K
- Likes41.8K
I don't get it... this is the correct way to read...
Kai Cenat kept having to look up what certain words meant during his reading journey to become a better speaker 😭 xitter.cf/TheClipBot/status/2011…
Uncle O'fear retweeted
Our CFO asked me to "audit" our software subscriptions last week.
He sent me a spreadsheet with 200 rows. Slack, Zoom, Jira, Notion, Trello, Monday, Tuesday, Wednesday.
He wanted me to survey the team to see which tools were essential.
I told him: "Surveys are for people who care about feelings. I care about OpEx."
I deleted the spreadsheet.
Instead, I logged into the corporate Amex portal and reported the card as lost.
Every single auto-renewal in the company failed instantly.
I call this "The Scream Test."
It’s simple Darwinian procurement.
If a tool goes down and nobody runs to my desk screaming within 4 hours? We didn't need it.
The Marketing team was at my door in 10 minutes begging for Adobe. We renewed it. The Sales team was crying about the CRM in 20 minutes. We renewed it.
But here’s the interesting part.
The HR department’s "Employee Wellness & Engagement Portal" ($12,000/year) has been down for six days.
Not a single person has noticed.
I didn't just save money. I quantified the exact value of our corporate culture.
It is zero.
Stop auditing. Start unplugging. If it’s important, they’ll scream. If they don't scream, it’s just noise.
Uncle O'fear retweeted
Apparently saying “adults should spend less than they make” will have you like this
This Post is from a suspended account. Learn more
Uncle O'fear retweeted
The funniest thing has officially happened to the International Association for Cryptologic Research.
Brother, they're out here reinventing CSVs.
TOON (Token-Oriented Object Notation) is out for some days now and it aims to make communication with LLMs more accurate and token-efficient.
The TOON topic is now one of the hottest news on the LLM market and it might actually matter.
𝗪𝗵𝘆 𝗜 𝘁𝗵𝗶𝗻𝗸 𝘀𝗼:
I was initially hesitant to cover this, potentially being another hype to quickly fade, but:
✅ The format has been shown to increase the accuracy of models while decreasing the token count. I was not sure if there were any accuracy retention studies made, it seems there were.
✅ Token efficiency is extremely important when working with Agentic Systems that require a lot of structured context inside of their reasoning chains. And we are moving towards a post-PoC world where there is a lot of emphasis placed on optimisation of the workflows.
𝗔 𝘀𝗵𝗼𝗿𝘁 𝘀𝘂𝗺𝗺𝗮𝗿𝘆:
- Token-efficient: typically 30-60% fewer tokens on large uniform arrays vs formatted JSON.
- LLM-friendly guardrails: explicit lengths and fields enable validation.
- Minimal syntax: removes redundant punctuation (braces, brackets, most quotes).
- Indentation-based structure: like YAML, uses whitespace instead of braces.
- Tabular arrays: declare keys once, stream data as rows.
An example:
𝘑𝘚𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵:
"shopping_cart": [
{ "id": "GDKVEG984", "name": "iPhone 15 Pro Max", "quantity": 2, "price": 1499.99, "category": "Electronics" },
{ "id": "GDKVEG985", "name": "Samsung Galaxy S24 Ultra", "quantity": 1, "price": 1299.99, "category": "Electronics" },
{ "id": "GDKVEG986", "name": "Apple Watch Series 9", "quantity": 1, "price": 199.99, "category": "Electronics" },
{ "id": "GDKVEG987", "name": "MacBook Pro 16-inch", "quantity": 1, "price": 2499.99, "category": "Electronics" }
]
}
𝘞𝘩𝘦𝘯 𝘦𝘯𝘤𝘰𝘥𝘦𝘥 𝘪𝘯𝘵𝘰 𝘛𝘖𝘖𝘕 𝘧𝘰𝘳𝘮𝘢𝘵:
shopping_cart: items[4]{id,name,quantity,price,category}:
GDKVEG984,iPhone 15 Pro Max,2,1499.99,Electronics
GDKVEG985,Samsung Galaxy S24 Ultra,1,1299.99,Electronics
GDKVEG986,Apple Watch Series 9,1,199.99,Electronics
GDKVEG987,MacBook Pro 16-inch,1,2499.99,Electronics
𝗥𝗲𝘀𝘂𝗹𝘁:
✅ 43% savings in token amount.
✅ Directly translates to 43% savings in token cost for this LLM input.
❗️ Be sure to know when NOT to use the format (and always test it for your application specifically):
- Deeply nested or non-uniform structures.
- Semi-uniform arrays.
- Pure tabular data.
ℹ️ I will be testing it in the upcoming weeks.
Let me know if you have already tested TOON and what are your takeaways! 👇
#LLM #AI #MachineLearning
Nancy Mace knowing that Jarvis calls her a snack will never not be funny
The internet works in mysterious ways.
Somewhere between trolling and political debates online, a group of strangers with deeply opposing views came together to help a young woman get brain surgery.
@damintoell, who is a lover of mayo, regularly mocks one of my superfans @jarvis_best. And from time to time we troll each other. So when Damin got suspended on X his followers let me know. Days later, the same group told me about a young woman @BootleggedChad needing brain surgery.
I called President Trump and the rest is history.
Follow these magnificent degenerates: @damintoell, @jarvis_best and @BootleggedChad.
The internet remains undefeated.
God is good.
HOLD THE LINE.
Uncle O'fear retweeted
Having children will also shift the vibe in your house.
Right now, you can grab a $5 roll of red tape and cover your house’s overhead lighting to entirely shift the vibe of your house.
This is the hack if you can’t replace your lights with Incandescent bulbs.
Uncle O'fear retweeted
People actually caught the satire. They just didn't think much of it because it was done so ineptly, so hamfistedly that it made the film endearing. (The same is true of Robocop.)
-The bugs weren't portrayed as sympathetic or misunderstood. Nothing in the film gave any indication they were anything but a murderous, hostile race.
-The worst you can say about the human forces is that they're incompetent in battle; for all the talk of fascism, none of the historical drawbacks of authoritarian societies were shown. The human leader stepped down and was replaced after the disaster on Klethandau instead of clinging to power. Military service was voluntary. Crucial decisions weren't delayed because generals were afraid to wake up Dear Leader. The humans didn't even start the war. (People insist that it was implied that they did, but this isn't actually in the movie.)
-Verhoeven wanted to make some biting satire about militaristic jingoism in the late 90s; American culture was neither militaristic nor jingoistic and he was satirizing a strawman.
Paul Verhoeven can make a good action movie but he's an artistic hack. The man compared Robocop to Jesus and thought Showgirls was a great flick.
The movie is dumb and people like it in spite of the director's intentions, which is funny, but you should really read the book.
Uncle O'fear retweeted
"This sidebar on counting C macro args needs to go deeper to help people understand."
3 weeks later...
"Here's why <10 lines of code to count at comp time requires advanced degree," My cost? 10k words and my sanity.
Recursive macros in C, demystified: h4x0r.org/big-mac-ro-attack/
Uncle O'fear retweeted
Saw a “sexy Curtis Sliwa” costume and I really think it’s time for the mayor’s race to be over
I really don't understand how someone could look at prime ribeye. Literally one of the most expensive cuts of meat and think this is some sort of gotcha. (Also yes this is from March so she's apparently just lying, but the point stands)
4 steaks...$110. Trump did this.
Readers added context they thought people might want to know
OP removed dates from prominent package. Righthand package shows dates March 4 and March 7 2025.
Donald Trump had only been in office for 6 weeks.
Cattle herds had been declining before he took office.
cnn.com/politics/live-…
feedlotmagazine.com/news/beef_news…
When you don't know what the Internet or AWS are.
If a company can break the entire internet, they are too big. Period.
It's time to break up Big Tech.
Readers added context they thought people might want to know
AWS is not a monopoly. It represents 30% of the web.
Sen. Warren also overlooks the fact that many providers went down because their services relied on a single AWS region, a malpractice that is not Amazon's fault.
It did not affect Google, X, and many others.
statista.com/chart/18819/wo…
x.com/gergelyorosz/s…
en.wikipedia.org/wiki/Redundanc…
docs.aws.amazon.com/whitepapers/la…
My hottest hot take is that if people were more litigious about copyright and open source was more litigious about licensing then we'd never end up with AI companies pretty much stealing everyone's data to create their models.
