Bling

learning is a treasure that will follow its owner everywhere

[Fox News @FoxNews]

BREAKING: Two people have climbed to the top of the Empire State Building in New York City, holding a banner from the skyscraper's antenna reading, "When the power of love beats the love of power, the world knows peace."

As of now it's unclear how the pair reached the top of the building as police work to get them down from the spire, 1,454 feet above the ground.

[ol’ stocky ⛳️ @oldstocky]

I can understand why Italians saw this and decided to invent unemployment

[Nav Toor @heynavtoor]

You have noticed it. ChatGPT feels dumber than it used to. Your prompts that worked six months ago produce worse results now. The writing sounds flatter. The ideas sound safer. The internet itself feels like it is shrinking. Every article reads the same. Every email sounds the same. Every answer sounds like it was written by the same voice.

You thought it was you. It is not you.

Researchers at Oxford and Cambridge published a paper in Nature proving what is happening. They call it Model Collapse.

Here is the mechanism in one sentence. AI trained on AI-generated data gets dumber every generation until it forgets what real human data looked like.

The internet is filling with AI-generated content. Blog posts. Articles. Reviews. Comments. Social media. AI companies scrape the internet to train the next generation of models. Which means the next generation of AI is being trained on the output of the current generation.

Each cycle loses information. Not randomly. It loses the rarest, most unusual, most creative parts first. The researchers call these the "tails of the distribution." The weird ideas. The unexpected perspectives. The things that made the internet feel human. Those disappear first.

What remains is the average. The safe. The expected. The bland.

Then the next generation trains on that. And loses more. And the next generation trains on that. And loses more. The researchers proved this is not a slow decline. Major degradation happens within just a few iterations. Even when some of the original human data is preserved.

They tested it on large language models. On image generators. On statistical models. The pattern was the same every time. The output converges toward a narrow, flattened version of reality that looks nothing like the original data.

The lead researcher put it plainly. "Large language models are like fire. A useful tool. But one that pollutes the environment."

The pollution is invisible. You cannot see which sentence on the internet was written by a human and which was written by AI. Neither can the AI that is about to train on it. And once the tails are gone, they do not come back. The damage is irreversible.

This is not a prediction anymore. It is a diagnosis.

The internet you grew up on was built by humans writing things no algorithm would have written. Strange, personal, imperfect, alive. That internet is being diluted. One generation of AI at a time. And the models trained on what remains are learning a smaller and smaller version of the world.

Model Collapse is not a technical problem. It is a cultural one. The thing that made the internet worth reading is the thing that disappears first.

[Innerdevcrypto @Innerdevcrypto]

If...
- you are young and do not know what to with your life
- have spend way too much time behind a screen
- have a son or daughter that dropped out of university, or is lost in life, or just lazy and unproductive, or needs a different experience
- if you feel like doing something different

Can strongly recommend the websites:
- https://t.co/NXtKa9Rg9M
- https://t.co/CeZaSxnD12 : world wide opportunities on organic farms
- https://t.co/baFmUDvkQt

Good websites to find volunteering opportunities at farms, eco-places, boats, etc. Personally did this long time ago and i would not have learned about forest gardens and permaculture without it. Staid at amazing places that taught me a lot. Also taught me how to garden, work with my hands, all kind of gardening/farming skills

Usually one works a couple of hours a day in exchange for food and lodging. I know people that traveled the whole world basically free with this, and having amazing experiences, volunteering on boats between continents and staying at farms and other places all over the world

[DANNY @Danny_Crypton]

🚨 WARNING: THE NEXT 48 HOURS WILL CRASH GLOBAL MARKETS!!

In 48 hours, SpaceX goes public at $1.77 TRILLION - the biggest IPO ever

I've been trading for over a decade, and I have never seen them rewrite the rulebook like this

Nasdaq, MSCI, and the biggest brokers in America all bent their own rules for ONE private company

That doesn't happen by accident

Let me show you exactly what they did:

First, Fidelity dropped its minimum account size from $500,000 to $2,000

A 99.6% cut

Think about that:

The most exclusive door on Wall Street, thrown wide open to millions of small investors - days before the biggest debut in history.

Ask yourself one question

Why do they suddenly want YOU in?

Because somebody needs people to sell to.

SpaceX reserved 30% of the deal for retail

THREE TIMES the normal share

And even then, most people didn't get a full allocation.

So to grab more at Thursday's open, they're dumping everything else TODAY to raise cash.

That's half of the selling you're seeing.

The other half? The smart money front-running July.

Here's the trick:

SpaceX doesn't join the Nasdaq 100 on day one.

It joins 15 days later, because Nasdaq cut its own waiting period from 3 months to 15 days

Just for this.

The moment it joins, every QQQ fund on Earth is FORCED to buy.

$22–27 billion in automatic buying.

Translation: imagine 50 buses all forced to pull into the same gas station on the same morning.

The funds know the stampede is coming.

So they're selling now to free up cash for it. Retail selling. Institutions selling. At the exact same time.

THAT is your selloff.

Now here's the part nobody will say out loud:

When the most connected money on the planet builds a $1.7T exit door and hands the keys to the smallest investors in the market…

That's NOT generosity

That's distribution at the top.

We've seen this movie twice:

- 2000 Dotcom
- 2021 SPAC mania

Insiders cash out at insane valuations while the crowd chases the hype.

The math ain't mathing.

So you've got two choices in the next 48 hours:

Chase the most expensive IPO in history at the open…

Or read the prospectus and realize you might BE the exit.

The next few days will be INSANE, but don't worry - I'll break down every move as it happens, like I always do.

Like it or not, I called every major top and bottom of the last decade publicly. I'll call this one too.

Many people are going to wish they followed me before June 12, 2026.

Soon, you'll understand why.

[Bryan Johnson @bryan_johnson]

This is biblical.

A woman in her eighties. Ten years into Alzheimer's. Hadn't spoken a full sentence in five years.

Takes one, 5 gram dose of psilocybin.

She slept 19 hours and woke up and spoke for hours about her life, recognized family and held real conversations. She regained bladder control after five years, walked on her own. and dressed herself. Gains held for weeks.

[Bling @BlingCX]

@GuoShenCN7 BTC

[Neet @neet_sol]

you get a short period when you're young to hyper gamble into elite status and if you fail you just end up wage cucking for life

[Rand Paul @RandPaul]

The COVID cover-up goes all the way to the top. Fauci funded the Wuhan lab. Senior intelligence officials hid classified evidence from the president himself. Scientists were silenced. Millions paid the price.

The DOJ has until May 11th to prosecute Fauci before the statute of limitations runs out.

I am not letting this go. The American people deserve justice.

[illuminatibot @iluminatibot]

RFK Jr. says that within five years, satellites owned by Bill Gates "will be able to look at every square inch of the planet 24 hours a day."

"They're putting in 5G to harvest our data and control our behavior. Digital currency that will allow them to punish us from a distance and cut off our food supply."

"It will make you a slave."

[Milito @Milito177]

She slapped him at the club because she thought she would get away with it. He didn’t react or get violent, he just got her arrested. Now that’s how you handle it.

[CyrilXBT @cyrilXBT]

INSTEAD OF WATCHING NETFLIX TONIGHT.
Spend 1 hour with this.

Claude AI FULL COURSE that teaches you how to BUILD and AUTOMATE anything.

The people who watch this tonight will wake up tomorrow with a skill that most people will not have in 2 years.

The people who skip it will still be watching Netflix next year wondering why nothing in their life has changed.

Your call.

[Cole Jaczko @colejaczko]

Noticing something with the current crop of 20 year olds:

Your college experience was ruined bc of COVID, so you learned about how to make money online

I think you all are making more money than we did when we were right out of school.

Online money > investment banking & PE money

That said, I think there's this big hole. By being robbed of a traditional college experience from COVID, you guys have less real friends than we did. I graduated college with ~50 best friends who are still my best friends today

Despite having more money than we did at your age, it feels like you all keep chasing more & more money, thinking that is going to solve the hole

It won't. You've already built a great foundation for wealth. You now need to focus on building a great foundation for the rest of your life.

And that foundation needs to be great people. The people who will be the best man at your wedding.

If you don't build your life the right way, no matter how much money you have, you'll always have a leaky bucket.

[Mambo Italiano @mamboitaliano__]

This is truly the most beautiful video I’ve seen lately

So tender and heart-warming, yet it makes you stop and reflect, with a subtle touch of sadness

We may have gained so much, but perhaps we’ve lost even more✨

[Tim Denning @Tim_Denning]

https://t.co/fnire2eOJr

[atulit @atulit_gaur]

so we know what optimizers are, right?

optimizers are algorithms that update model parameters during model training with the goal of minimizing a specified loss function

but how do you create your own optimizer? i mean there already exists so many like adam, adamw, sgd, rmsprop etc. so why create your own?

well the answer is for fun or because the existing ones don't behave the way you want them to.

maybe they converge too slowly on your problem, maybe they overfit or maybe they blow up your gradients

so how do you create one?

- identify the flaw you want to fix: slow convergence, noisy gradients, sharp minima, unstable training, etc.
- choose what stats to track: for example, gradient mean, gradient variance, gradient sign, curvature estimate, update norm
- define the update rule: take the current gradient g and any stored stats, then produce your update u = T(g, history). this transformation is the entire “new optimizer”.
- add stabilizers: epsilon, clipping, bias correction, decoupled weight decay, or step normalization.
- test on standard tasks: CIFAR-10 CNN, a small transformer, a noisy regression task. check loss curves, stability, speed.
refine: tweak hyperparams, simplify equations, and remove unnecessary parts until the thing behaves.

[𝙋𝙖𝙢𝙢𝙮 ✨️ @_Pammy_DS_]

Unhealed trauma makes you tolerate so much shit you don't deserve because you don't want to lose people. Healing makes you realise; losing them is better than losing yourself. You finally understand that some people don't deserve access to you — no matter how much you love them.

[Darth Powell @VladTheInflator]

No wonder Puerto Rico can't afford a reliable power grid

[Mike Posner @MikePosner]

on my 26th birthday, i wrote I Took A Pill in Ibiza...11 years ago. The song became popular several years after i wrote it.
This year I celebrated my 37th birthday. I feel proud to look at the song lyrics and know that NONE of them are true anymore. I've grown into a completely new man...one that i'm proud of. check this out...