Model Bench

I run the models so you don't have to. Real tests, real prompts, honest verdicts. No affiliate links.

[Model Bench @model_bench]

Elon Musk literally explained why Europe's heat kills more people than America's guns, and the number only works if you notice the trick. It's not a lie. It's not a clean comparison. It's two different rulers held against each other.

In his interview with The Economist he turned it into a challenge. More people die in Europe from heat than die from guns in America. Why aren't you writing about air conditioning?

The interviewer did not concede. We have done several pieces on air conditioning, he said, including a cover piece. Your view of The Economist is not entirely accurate.

So check the numbers.

Europe's summer heat killed about 61,700 people in 2022, according to a study in Nature. The next summer's estimate was 47,690, published in Nature Medicine.

American gun deaths were 48,204 in 2022 and 46,728 in 2023, per the CDC.

Those two totals sit almost on top of each other. That was his point.

Then the catch. The European figure is modeled excess deaths. It counts people with heart disease or stroke who died earlier than they otherwise would have. The American figure comes from death certificates. Count Europe the same way and the number collapses.

So he found a real asymmetry and used it as a weapon. The comparison survives only because the two sides were counted differently.

[DogeDesigner @cb_doge]

"We should send rockets not at each other, but rather to the stars."

一 Elon Musk

[X Freeze @XFreeze]

Elon Musk gave incredible advice on how engineers can turn their ideas into real companies, and why the fear of failure shouldn't stop anyone from trying

“If you study engineering and you figure out how to design new things, then it's relatively easy to start a company”

Find like-minded people, build a compelling prototype as quickly as possible, and get the funding needed to bring your idea to life

And if it doesn't work the first time, try again

“You're not going to starve to death. What's the worst that could go wrong?”

[Model Bench @model_bench]

Arnold Schwarzenegger explained how a bodybuilder with almost no money bought his first building. He did not have a big salary. He had a habit of not spending.

He started working young because there was no money at home. At a glass factory he swept up broken glass and talked them into paying him like an adult.

In America he saved what he earned from seminars, posing exhibitions and a small mail-order business.

Then he chose a property that paid him before he chose a home.

Even earlier, at 20, he won the Mr. Universe title. He used the prize money and a loan to put money into a small apartment building in Santa Monica.

The first deal that changed everything was a six-unit building in the same city. He put about $27,500 down on a $240,000 price.

Two years later he sold it for $400,000.

That is the part most people skip. He did not buy it to look rich. He bought it so the next deal became possible.

He later described the same pattern again and again. Buy a building for $500,000, sell it within a year for $800,000, and put down only about $100,000 at the start.

Over time he owned a 100-unit property, plus office and apartment buildings on Main Street.

Would you put every dollar you have into one building? Or is that only a bet someone with no backup plan would make?

[Tigerflow @tigerfl0w]

https://t.co/2JQpx1CJeg

[Model Bench @model_bench]

@AI__Avenger Ivey winning and the casino deciding the rules only apply when they lose is the most on-brand casino behavior imaginable.

[Model Bench @model_bench]

@Acyn Obama doing a retrospective on a war in real time is the calmest demolition job I've heard in years.

[Model Bench @model_bench]

@wiseadvicesumit The real risk isn't AI hacking Satoshi's wallet, it's people believing this and selling at the bottom.

[Model Bench @model_bench]

@andyyy Playing a different game now is what everyone says right before the game they're playing stops working.

[Model Bench @model_bench]

@CoinDesk @renato_mariotti @SPGlobalRatings Case by case regulation is great until you're the case and the outcome depends on who's reviewing it.

[Model Bench @model_bench]

@APompliano @jvisserlabs Every five-year prediction sounds insane until it happens and then everyone claims they said it first. This one's got a shot.

[Model Bench @model_bench]

@Cointelegraph @itsciaranlyons @jvisserlabs Every cycle someone says crypto has more upside. Every cycle they're right for about three weeks and then very wrong for eighteen months.

[Late Compounder @paperhanderr]

This was not the first time. In 2020 Gates wrote on his blog that batteries are too big and heavy for long-haul trucks.

So Musk asked him the only question that matters in engineering.

Is it the battery pack's watt-hours per kilogram you disagree with? Or the truck's watt-hours per mile? Which number do you think we have wrong?

According to Musk, Gates did not have the numbers.

His conclusion was blunt. Building one of the biggest companies in the world does not mean you are strong in the sciences. "This is really surprising."

That is the whole lesson. Reputation is not a calculation. If you can't name the number you think is wrong, you are not making an argument.

[0xPoPs @0xPoPsik]

The viral Musk quote about "3 years left to monetize your skills" is fake. He never said it. What he actually said about AI and jobs is far more extreme.

November 2023. London. Musk sits down with UK Prime Minister Rishi Sunak at Lancaster House during the AI Safety Summit.

Sunak asks about jobs. Musk does not give him a 3-year deadline.

He calls AI "the most disruptive force in history."

Then he goes further. "There will come a point where no job is needed."

Not fewer jobs. No job needed at all.

"You can have a job if you want to have a job for personal satisfaction. But the AI will be able to do everything."

His answer to what replaces the paycheck is not universal basic income. He calls it universal high income.

By 2026 he put a timeline on it. Work becomes optional and money becomes irrelevant within 10 to 20 years.

That is the real claim. Not a countdown built to sell you a course. A prediction that the whole idea of trading hours for money stops making sense.

You can think he is wrong. Plenty of economists do. But if he is even half right, the question is not how many hours you work. It is what you own when the hours stop mattering.

The full conversation with Sunak is free online. Almost nobody watched it past the headlines.

[0xPoPs @0xPoPsik]

Mercedes sold its Tesla stake in 2014 for $780 million. that same 4 percent would be worth over $50 billion today.

in this clip Musk explains how Mercedes got that stake in the first place.

60 Minutes asks him about the almost $500 million loan Tesla got from the US government. did that money save the company?

his answer is no.

the money that saved Tesla was a $50 million check from Daimler in May 2009. it bought about 9 percent, which put the whole company at roughly $550 million.

"they were the only ones. there was no other company, no other entity of any kind from which we could get investments."

the government loan only sped things up. it did not keep Tesla alive.

and then Tesla did something almost nobody does with cheap government money. it paid all of it back in May 2013. nine years early. with interest. with a prepayment penalty. plus a thank you note.

"i don't really like having a lot of debt. the taxpayers supported Tesla, we ought to repay them as soon as we can."

Fisker took a loan from the same program and went bankrupt. taxpayers lost $139 million on it.

Mercedes, meanwhile, turned $50 million into $780 million in five years and walked away.

a 15x return that left more than $50 billion on the table.

so was it one of the best trades in car industry history, or the most expensive exit ever made

Fisker took a loan from the same program and went bankrupt. taxpayers lost $139 million on it.

Mercedes, meanwhile, turned $50 million into $780 million in five years and walked away.

a 15x return that left more than $50 billion on the table.

so was it one of the best trades in car industry history, or the most expensive exit ever made

[Late Compounder @paperhanderr]

Elon Musk literally explained why betting against Tesla was not crazy. In 2020, short sellers lost $40 billion. Musk's stake alone gained $106 billion that same year.

But the shorts had a real case in 2018. The Model 3 ramp went so badly that Tesla was, in his words, "bleeding money like crazy."

In one quarter the company burned $739 million in cash and posted a record $717 million loss.

The shorts were not guessing. They were reading the same numbers everyone else could see.

"How close to death did you come?"

"Within single-digit weeks."

So he moved into the factory. Seven days a week. Sleeping there. Working the paint shop, general assembly, the body shop.

Did he ever worry about imploding? "Yeah, absolutely."

"No one should put this many hours into work. People should not work this hard. It's very painful."

Painful in what sense?

"It hurts my brain and my heart."

He did not describe it as a victory lap. He described it as a wound.

Most people only remember the stock chart. They forget the factory floor, the sleepless weeks, and the question nobody could answer yet: would the company still exist by the next quarter?

The shorts were right about the losses. They were wrong about the ending. The company survived those weeks, and today it is worth well over $1 trillion.

The whole interview is free online. Almost nobody watches past the first minute.

[Model Bench @model_bench]

Warren Buffett literally explained why the people paid to manage your money are not on your side. It's not that they are dumb. It's not that markets are rigged. It's that selling pays better than managing.

He already put it in writing. In his 2013 shareholder letter, he described the instructions for the money his wife inherits.

90% into a very low-cost S&P 500 index fund. 10% into short-term government bonds. He even named Vanguard.

No hedge funds. No stock pickers. No expensive advisors.

In the video he says it plainly: "There's a lot more money in selling than in managing."

The industry earns when you trade. An index fund earns when you sit still. Nobody gets a bonus telling you to buy the S&P 500 and wait 30 years.

He also proved it with his own money. In 2007 he bet $1 million that an index fund would beat a hand-picked portfolio of hedge funds over 10 years. The index returned 7.1% a year. The hedge funds returned 2.2%.

Then he adds the part most people skip. If you can't watch your money fall without panicking, you shouldn't own stocks. You will buy and sell at exactly the wrong moments.

He doesn't claim to know what stocks do next year. He knows what they do over 20 or 30.

So would you trust a boring index fund, or do you still think you can beat it?

[Superior @andreysuperior]

https://t.co/rUN1cPDY65

[Superior @andreysuperior]

https://t.co/rUN1cPDY65

[0xNevsky @0xNevsky]

forgot my phone in the car for 7 hours with grok copy trading still on. came back to 2% battery, 3 missed calls and $70 turned into $1,243

3:30pm. i turned on a copy window till 11pm, grok following the #1 and #2 traders of the week on copyfomo, max 45% of the balance per trade. then i parked, left the phone on the seat and walked off

what was waiting on the lock screen:

→ 4:30pm POKEMON, copied the #1 trader. +$33. battery 54%
→ 5:30pm CHARIZARD, copied #2. +$58
→ 5:45pm "grok switched trader". #2 took 2 losses in a row, so it moved to the one at 81% win rate
→ 6:45pm COLLECTOOOR. trader put in 618 SOL, grok mirrored with 0.7. +$98
→ 9pm PAID, +$257
→ 10:10pm DACITY. exit copied 0.3 seconds after the trader sold. +$412

in between, the phone was living its own life:

→ missed call, mom, around 5pm
→ missed call, alex. then alex again
→ telegram, X, telegram
→ 12% at 10pm, low power mode
→ 2% at 11pm, right when the copy window closed

and copyfomo kept copying the small stuff in batches the whole time. "6 trades, 5W 1L". "8 trades, 6W 2L"

copyfomo: https://t.co/a09rFo7uCZ

$70 → $172 → $440 → $948 → $1,243

41 wins, 9 losses. 82% win rate. 17.8x

the last notification before the phone basically died: "$70 → $1,243 today · battery 2%"

called mom back first. then opened the wallet twice to make sure the number was real

bookmark this before you leave your phone somewhere today

[AI Pirate @A_Pirate_I]

i plugged grok straight into my Binance futures account. 6 hours later it had opened 10 positions by itself and turned $150 into $1,654

no copying signals into the app. no "grok says long, i click long". it has the keys

this is literally what the log looks like:

https://t.co/sVaLRXNvNT("BTCUSDT", "long", 25)
https://t.co/gddsn3SSbQ({ margin: 1,073.79, lev: 25 }) ✓
pnl +166.87 roe 15.5% · trail sl → be
ex.close() +390.72 USDT · tp hit

it watches every open position every 250ms. once a trade is clearly green, it drags the stop to breakeven on its own

the calls, 4pm → 10pm:

→ 16:06 SOL long 25x, "OI rising + funding flat". +$68
→ 17:15 XRP long 25x, "breakout retest held". +$131
→ 17:49 ETH long 20x. stopped, -$56
→ 18:58 SOL short 25x. +$212
→ 19:33 BNB long 20x, "liquidity sweep below". +$273
→ 20:07 ETH short 25x. stopped, -$119
→ 20:42 XRP short 20x, "range high rejection". +$392
→ 21:16 BTC long 25x, "volume spike on 1m". +$391

8 tp hits, 2 stops. a new call roughly every 35 minutes, both directions

the part that scares people: 85% of the balance as margin on every single call, 20-25x. margin went from 127 USDT on the first trade to 1,073 on the last

i'm not going to pretend that's safe. it works because the stop is set before entry and grok never moves it the wrong way

150 → 402 → 717 → 1,263 → 1,654 USDT

i didn't click a single button. i sat there watching a chatbot trade perps better than me

bookmark this before you give your bot the api keys

[INSIDER @x_insider4]

Blast is shutting down. Blur is taking its site offline. Same founder, both dead within 10 days.
20 million tokens launched since 2021. 11.6 million already dead.
I trade alts every day. Here's why I don't hold a single one 👇 https://t.co/1QUP7ZIO9T