@mikealfred Mike, you should watch the family man with Nicolas Cage.
English
Samuel Ohev-Zion
299 posts
Imagine you had to choose your life at age 40:
Option A:
Single. No kids.
$50M net worth.
850 credit score.
Private jet.
Option B:
Married. 2 kids.
$3M net worth.
Drive a Toyota.
10 BTC in cold storage.
500 credit score.
Fly Southwest.
Be honest, which life are you choosing?
English
@Darren12555 @mikealfred @zachrudolph87 Some wealthy people tend not to care about paying bills on time and/or they live volatile lives. While credit score might mean a lot to the average person, it doesn't matter in the lives of others. Not an end all be all.
English
@mikealfred @zachrudolph87 Generally if you have 3 mil in assets available to you, you’d have enough to keep your credit in good shape
English
Four B-2 stealth bombers carried out a successful operation in Iran tonight and are now en route back to Whiteman Air Force Base in Missouri.
English
Samuel Ohev-Zion 리트윗함
🚨 WOW! Sen. Mike Lee just went WAR MODE on the floor for the SAVE America Act
"Not a CHANCE IN HELL am I giving up on this!"
"We're NOT going away. I'M not going away. I will be back on this floor, in this chamber, DAY AFTER DAY, WEEK AFTER WEEK, MONTH AFTER MONTH, as long as it takes. I'm. Not. Giving. UP."
"This is about the American dream, this is about the fact when you allow noncitizens to vote, you are ROBBING something from the American people!"
"Something which is distinctively, rightfully, THEIRS."
"We MUST not tolerate it — and WE WILL NOT."
English
@zerohedge Foolish to think that the hard-line psycho Iranians will ever agree to anything. They'll die before ever surrendering.
English
So they are negotiating after all
zerohedge@zerohedge
IRAN REJECTS U.S. PROPOSAL DELIVERED VIA MEDIATOR, VOWS TO CONTINUE FIGHTING
English
Market stupidity same as Deepseek BS.
Efficiency gains like TurboQuant often lead to more overall usage rather than less demand — a classic Jevons paradox effect seen repeatedly in AI (e.g., after other compression/optimization breakthroughs):Cheaper/faster inference (lower cost per token, viable longer contexts) unlocks new applications, higher adoption, and more queries. This increases total AI workloads, which still require memory (weights, activations, system DRAM in servers/edge devices).
It enables edge AI growth: Better on-device performance on existing smartphone hardware could accelerate local AI features without proportional hardware upgrades, but widespread adoption might drive demand for higher-spec phones over time.
Analysts largely view it as long-term positive/neutral for memory demand: hyperscalers get better ROI, leading to more infrastructure spend overall. Short-term volatility in stocks doesn't change the structural supply tightness or pricing of memory.
English
Efficiency gains like TurboQuant often lead to more overall usage rather than less demand — a classic Jevons paradox effect seen repeatedly in AI (e.g., after other compression/optimization breakthroughs):Cheaper/faster inference (lower cost per token, viable longer contexts) unlocks new applications, higher adoption, and more queries. This increases total AI workloads, which still require memory (weights, activations, system DRAM in servers/edge devices).
It enables edge AI growth: Better on-device performance on existing smartphone hardware could accelerate local AI features without proportional hardware upgrades, but widespread adoption might drive demand for higher-spec phones over time.
Analysts largely view it as long-term positive/neutral for memory demand: hyperscalers get better ROI, leading to more infrastructure spend overall. Short-term volatility in stocks doesn't change the structural supply tightness or pricing of memory.
English
$MU $SNDK Dips like this don't come often.
Deepseek was BS
TurboQuant is another BS.
Act accordingly.
English
From Grok:
TurboQuant compresses KV cache memory to ~3 bits per value (6x reduction) with zero accuracy loss and negligible runtime overhead—Google calls it "lightweight" and "accelerator-friendly," with near-zero quantization time.
The 8x speedup is from slashing memory bandwidth bottlenecks in attention, not ramping chip clocks or utilization. No added stress on compute units; memory access (a big power hog) drops sharply.
No data supports shortened chip lifespan (7→3 years) or "immense stress"—GPUs are built for sustained high loads, and this enables cheaper hardware overall. Energy/cooling per inference falls due to efficiency; total bills trend down despite higher throughput potential.
Early ports (e.g., Apple Silicon) confirm easy drop-in with net gains.
English
$MU $SNDK This is actually pretty bad
Google TurboQuant destroys your chips 57% faster due to immense stress on Compute. Cuts lifespan from 7 years to 3. Then add liquid cooling and energy bills.
Free lunch. Right. Gemini 3 did the math 👇
Trade Whisperer@TradexWhisperer
$MU $SNDK Let me expose the REAL downsides of Google TurboQuant Compression for you: Google's TurboQuant has big downsides. It uses more GPU power and energy. The bill shows up in four places: power, latency, logic, and quotas. GPU Power. The compression doesn't eliminate compute, it shifts it. Your GPU has to do extra work to pack and unpack vectors on every single read. Net result: higher energy consumption per inference, not lower. Latency. Saving VRAM sounds great until you see the Time to First Token spike. Unlike raw 16-bit memory where the GPU just reads data, TurboQuant requires mathematical transformations before the model can generate a single word. Gemini 3.1 Pro users are already reporting 60-90 second thinking loops on simple queries. It's not stuck. It's decompressing. Streaming. Text no longer flows word by word. Decompression happens in chunks. The UI freezes for 10 seconds, then dumps 200 words at once. Conversational feel is gone. Logic. TurboQuant is lossless for retrieval, not for reasoning. In complex coding or math, 3-bit quantization loses the subtle data spikes that represent edge-case logic. Great at summarizing a 100-page PDF. Hallucinate a library name in a Python script because that token's precision was compressed away. Quotas. Deep Thinking mode consumes more internal compute cycles, not less. Users are hitting quota limits after three messages. Some are getting 24-hour lockouts. No free lunch.
English
Google just dropped a compression algorithm that makes AI 8x faster while using 6x less memory. Zero accuracy loss.
It's called TurboQuant. Here's why it matters in plain English:
Every time you talk to ChatGPT or any AI, the model has to remember everything you've said in the conversation. That memory is called the "key-value cache." The longer the conversation, the bigger the cache, the more expensive it gets to run.
This is the single biggest bottleneck in AI right now. A 128,000-word conversation on a large model eats 40GB of GPU memory just for that one user. Scale that to thousands of users and you're burning millions in compute costs reprocessing the same data over and over.
TurboQuant compresses that memory down to just 3 bits per value (from 32 bits) without losing any quality. Independent developers tested it within hours and got exact matches against full-precision output.
What this actually means:
- AI models that needed a $10,000 workstation could now run on a MacBook
- Always-on AI agents become dramatically cheaper to operate
- Open-source models that were too big for consumer hardware suddenly fit
- The cost curve for every company running AI infrastructure just shifted
Developers are already porting it to Apple Silicon. No retraining required. It drops into existing AI systems without modification.
The AI cost problem isn't being solved by building bigger data centers. It's being solved by mathematicians figuring out how to do more with less.
English
@dialecticjawn @TheLongInvest Gaza leaders could have turned the place into a mini Dubai and improved the lives of their people and lived in peace with their neighbors. Instead they chose violence. Please blame the Hamas leadership.
English
@Sammy_CT @TheLongInvest Perhaps. But 85% of Israels support the ethnic cleansing of Gaza, over 40% support killing them en masse, and 100% live on stolen land. Jewish supremacism is a disease afflicting the entire culture.
English
@TheLongInvest Hard to fight wars at home when democrats block everything including fair elections which is how they get elected.
English
BREAKING: Trump: we have won the war.
Whatever you want to call it.
End it and fight your own wars at home.
English
Three themes. Twelve names. That's my 2026 list.
Chips: $MU $TSM $SNDK
Optics: $GLW $LITE $AXTI $AAOI $TSEM
Packaging: $AMKR $BESI $AMAT $LRCX
Trade Whisperer@TradexWhisperer
@Dixitagg When memory booms, usually the equipments follow and the market did exactly that already, very impatient. Packaging will boom as well. Personally, I am going after Fiber optics $GLW $LITE $TSM.
English
The Micron short thesis, in a nutshell
$MU closed 10% off the highs despite blowout earnings... the market is telling you the end is near for the memory super cycle
Steve Burns@SJosephBurns
“If all the news is great and the stock is not acting well, GET OUT -- which is a pretty simple thing most analysts don't know.” — Stanley Druckenmiller
English
Jury found Musk liable for securities fraud on 2 specific 2022 tweets (one saying the deal was "temporarily on hold" over fake accounts, dropping stock ~10%). Not liable on broader "scheme" claim or 1 other statement (podcast).
Damages are for class of shareholders who sold TWTR shares during the alleged artificial dip (May-Oct 2022)—out-of-pocket losses from depressed price. Jury set ~$3-8 per share per affected day; total likely billions once finalized by experts/formulas. Musk's eventual full $44B payment doesn't factor in, as this is separate harm to interim sellers.
He can appeal (standard for civil verdicts; to 9th Circuit). Not over. Next: post-trial motions, then likely years of appeals.
Sources: AP, Daily Journal court reports.
English
U.S. JURY FINDS ELON MUSK LIABLE TO TWITTER SHAREHOLDERS IN SECURITIES FRAUD LAWSUIT RELATED TO $44 BILLION TAKEOVER, COURT RECORDS SHOW
English
$MU PEG Ratio of 0.2
Is the market giving you a gift?
English
Samuel Ohev-Zion 리트윗함
$MU Massive earnings beat and stock is down 1.5%
TrendSpider@TrendSpider
$MU MASSIVE DOUBLE BEAT ~EPS: $12.20 vs $9.21 est ~REV: $23.86B vs $19.93B est 🟩 +1.89%
English
❗️❗️❗️URGENTE, DENUNCIA❗️❗️❗️
Esta es la desgracia de el pueblo de Cuba bajo el comunismo cubano, Un trasporte de alimentos por accidente dejó caer parte de su carga de huevos, Los cubanos de apie, el pueblo, hace un gran tumulto para tratar de salvar algunos huevos para poder comer.
❗️❗️❗️URGENTE, DENUNCIA❗️❗️❗️
Español
@TradexWhisperer 80% gross margin forecast is higher than Nvidia. Insane that anyone is selling. Market is just weak atm
English
Who sells $MU after a print like that?
Someone who was never really long to begin with. Event trader. Fear trader. Headline trader.
Fundamentals don't expire overnight. Patience is my edge.
English