A $1 unit of AI capability today falls to roughly 8 cents a year later if the current curve holds.
47% cheaper every quarter means fixed capability is basically half-price every ~100 days.
The hardware underneath is moving nowhere near that fast.
AI chip performance/$ has improved ~49% a year since 2023. quality-adjusted model capability is getting ~13x cheaper.
So most of the compression is happening above the silicon: better architectures, training, quantization, serving + brutal lab competition.
And cheaper intelligence gets consumed almost immediately.
@OpenRouter cut Luna pricing ~90% for

League of Legends World Championship Finals
Fight until there are no regrets!
@Polymarket
You are about to see the most results-obsessed 😡
and publicly pressured esports competition in the world 🤯
Here, 0-3 is called a spinning top 🌀
1-3 is called essentially 4-0 🤣
2-3 is called losing after going the distance 😭
At the Round of 16, solo songs are everywhere 🎵
The quarterfinals are called wiping sweat from your neck 🥵
The semifinals are called swimming home by yourself 🏊
The runner-up is called a lucky Round of 16 team 🥈
The champion is called a tailor-made cham

THE GRID THAT NEVER BLURS
Centuries ago, mosaic makers laid single colored stones into open grids.
From a distance you saw a face or a star, up close you saw only the individual stones, each with one color and one exact place.
The whole picture only existed because the small pieces agreed to stay where they were put.
Here comes @Synthesis_ETH , the 3333 stones are pixels and the grid lives on Ethereum which can be acquired freely.
Every piece starts with the same three things: a visible grid, a seed number, and a glowing core.
No two seeds produce the same result. Nothing is smoothed or h

$RAY RAY/USDT – Bullish Divergence + 3D Alligator Retest
RAY is showing a potential bullish reversal setup, with bullish divergence on the 4H timeframe aligning with a possible retest and bounce from the Alligator on the 3D chart.
Key Technical Points:
4H Bullish Divergence: Momentum is showing signs of recovery, suggesting selling pressure may be weakening.
3D Alligator Retest: Price is testing the higher-timeframe Alligator area, where a bullish bounce could support trend continuation.
Potential Reversal: A confirmed bullish reversal could bring the previous highs back into focus.
Upside S

It’s back above 82k. But if you think this is “reclaiming lost ground,” you’re giving this rebound far too much credit. #比特币短线站上8.2万美元
This rebound from $79,694 was built not on spot buying, but on the corpses of longs. That wick on the 15th sent $929 million in leveraged positions across the market into liquidation, with longs accounting for $777 million of that. After so many bulls were forcibly carried off the field, selling pressure temporarily dried up, and the price was lifted back by short covering and technical buying. Put simply, 82k is an “empty spot” left behind after leverage was

🚨 $NVDA: Price Action is cooking a brutal setup! Don't blink!
Chart Update.
Look at the structure on $NVDA. Bears are trying hard to push it down, but we are entering a major accumulation zone. Price action looks clean right here. The order book is heavy near the bottom as buyers are waiting like hungry wolves.
Current price is sitting around $214.13.
Immediate major support is rock solid at $209.00. This is the exact demand zone where whales are buying every single dip. If we hold this block, shorts will get absolutely squeezed.
On the flip side, the major resistance line sits at $228.00.

@BrandonBiggs247

U.S. Stock Premarket Analysis: Memory Chips Surge Across the Board, Tech Sector Starts to Rebound🔥
The Nasdaq is up over 1% in premarket trading today, NVIDIA has turned positive, up 1%, the AI sector is following suit, nonferrous metals have turned positive, and oil prices have pulled back
Large-cap tech stocks are generally posting solid gains. International oil prices are deeply oversold, Treasury yields are also falling, and Japanese and South Korean markets closed higher
Crypto Bro is still fairly bullish on tech stocks tonight. After all, it’s Monday’s opening, and Japan and South Korea

Gate ETF Newly Listed: $QNT
🔹 Trading pairs: #QNT3L / $USDT, #QNT3S / $USDT
🔹 Trading time: September 28, 2026, 17:00 (UTC+8)
🔹 supports 3x leveraged long and short trading for greater flexibility
Trade:
https://www.gate.com/trade/QNT3L_USDT
https://www.gate.com/trade/QNT3S_USDT
More: https://www.gate.com/announcements/article/101981

Which celebrities look kind just from their faces?

📈 U.S. stocks this week: It’s the tech sector’s turn to “show its work” 🍿
Tesla and Meta earnings + AI / memory chips
Can this round of results give tech stocks a boost?
Which company’s analysis are you most looking forward to?
Name it in the comments, and we’ll bring your questions to the host 🔴
Save the weekly schedule. See you in the livestream.
https://www.gate.com/live
‍#GateLive #美股 #TSLA #META

#Gate广场新人季 🧧 Get a red packet with your first post, up to 10 USDT! Gate Square’s Newcomer Posting Season is here!
New users are guaranteed a red packet with their first post. Join the leaderboard to share 15,000 USDT and exclusive Gate merchandise!
Three event perks
1️⃣ First-post reward: 100% chance to win, including large red packets worth 10 USDT and 3 USDT
2️⃣ Daily Quality Leaderboard: 3 winners each day, each receiving a 100 USDT trial voucher + official engagement
3️⃣ Monthly Creator Leaderboard: Share 10,000 USDT and Gate T-shirts/umbrellas/water bottles
✍️ Share your market views, tr

Instructor looks so awesome from this angle.

pov: you can't stop when this is the goal:
gm ct

#KOSPIFalls3.17%AtOpen 📉🇰🇷 KOSPI Drops 3.17% at Market Open!
South Korea’s stock market faces a sharp sell-off as the KOSPI falls 3.17% at the opening bell. 🔻 Investor sentiment is under pressure, and market participants are watching closely for the next move.
📊 Market volatility rises
🌏 Global economic sentiment remains in focus
⚠️ Investors reassess risk and market exposure
Will the KOSPI recover during the session, or is more downside ahead? 👀

Many trading strategies sound very logical—until you actually run them against historical data.
“Buy after a big drop.”
“Chase a breakout.”
“Enter when a certain indicator forms a golden cross.”
These rules all make perfect sense in your head, but what really determines their value is not how reasonable they sound, but what they actually earned and lost across different past market conditions, and how large the maximum drawdown was.
That’s why I increasingly believe backtesting isn’t exclusive to quantitative traders, but is one of the most direct ways for ordinary traders to validate their id

5 things to check before depositing in an ai trading vault.
never look at apy first. it's the last thing you should look at.
i'll use apex ms on @Infinit_Labs as the example.
1/ who is trading your money?
"ai trading vault" doesn't mean nobody's in charge. someone built the model, set the limits, and decides what happens when it breaks.
you want real people with a real track record.
apex ms is run by apex quantitative capital: ~50 years of combined experience across global banks, buy-side funds, crypto trading firms and exchanges. the team and its beneficial owners went through identity checks