TrueNorth

@dampedspring Curious what your second word was. Or maybe you started with something new.

Wordle 1,741 3/6* ⬜⬜🟨⬜🟨 🟩🟨⬜🟨🟩 🟩🟩🟩🟩🟩

@dampedspring Thank you

I wrote a little blog on Commodities in an investment portfolio. Some fun anecdotes and a primer 101 level analysis. Sign up for my blog its free and free of AI slop and always will be. Don't pay even if they try to make you. Link in comment.

@TradexWhisperer TY very much.

@TradexWhisperer What is the likeliest successor memory/storage medium? SOT-MRAM? using Beta phase tungsten? or something different?

$MU I worked 21 years as an HBM, DRAM & NAND engineer. AMA is open. Ask me anything. I'll drop rare insights where I can.

@F1 Everyone is over 7 seconds slower than Hamilton's 2020 qually lap. What a disaster this sport has become.

Ferrari's Charles Leclerc sets the Thursday pace in Bahrain, ahead of McLaren's Lando Norris and Haas' Ollie Bearman 🙌 Read the full report from Day 2 here 👉 f1.com/2026-Testing-T… #F1 #F1Testing

@nnzp1730 Roy Clark watching another guitarist like that, tells you all you need to know.

Fantastic…..watch to see that guitar work…..about half way in

@hannahapexfit APEX

LOSE FAT AND BUILD MUSCLE I built the Ultimate Guide which shows you exactly how to torch fat and build muscle before this summer. Like + comment "APEX" & I'll send you a FREE copy. (Taking it down in 24hrs only, must be following me to receive it)

@dampedspring Horrible. Didn't even finish it.

F1 is an absolutely terrible movie. Best picture nominee? Oof

@RyeNotBerben 113 people (including myself) thinking..."god I hope it's not me".

Why is it the people with the most unsolicited advice have the worst ideas?

@rubicon59 "new NAND memory"... HBF?

CPUs soon the new NAND memory? $INTC

@anilananth This is what will happen. A stateful AI capable of constructing and maintaining a "information/knowledge taxonomy" database, would improve query response immensely and dramatically reduce the need for processing resources. Most of which would be pushed to the edge.

We want LLMs to be factual. But #LLMs don't look up facts from some table, they recompute known patterns and knowledge for each query, thus wasting resources and making mistakes. What if specialized neural network layers could learn to store and retrieve factual knowledge in a way that mimics looking up information, and thus free up the Transformer layers for reasoning? @deepseek_ai seems to have done it! Here's my Substack post explaining DeepSeek's possibly potent change to the LLM architecture: wheremachinesthink.substack.com/p/deepseek-dro…

@paurooteri Hynix has promised to increase production 8x this year. Samsung 4x this year and MU doubling this year through yield/node advances and 2x again in 2027 (new clean room). I hope your Note that addresses this. TY

ありがとうございます 一昨年からずっと供給が足りなくなると主張しておりまして、最近実際の経済に影響が出て来ました 今のDRAMは簡単に増産出来る状況でもないので、供給は急には増えません メンバーシップでは詳細まで説明していますっ

@MichaelAArouet the " 'unregulated' sprint".... says it all

If he had lived earlier, he would have pushed for a moratorium on car manufacturing to give horse carriages a chance to catch up.

@RyeNotBerben And 7.5% of SpaceX

@RyeNotBerben Looks like a good call on this one.

@AlexanderFYoung Send

🔥 MIT just exposed every top AI model and it’s not pretty. They built a new test called WorldTest to see if AI actually understands the world… and the results are brutal. It doesn’t just check how well a model predicts the next frame or maximizes reward it tests whether it can build an internal model of reality and use it to handle new situations. They built AutumnBench 43 interactive worlds, 129 tasks where AIs must: • Predict hidden parts of the world (masked-frame prediction) • Plan sequences of actions to reach a goal • Detect when the environment’s rules suddenly change Then they tested 517 humans vs. Claude, Gemini 2.5 Pro, and o3. Humans crushed every model. Even massive compute scaling barely helped. The takeaway is wild.. today’s AIs don’t understand environments; they just pattern-match inside them. They don’t explore strategically, revise beliefs, or run experiments like humans do. WorldTest might be the first benchmark that actually measures understanding, not memorization. The gap it reveals isn’t small it’s the next grand challenge in AI cognition. (Comment “Send” I’ll DM you the paper)

@RyeNotBerben Not sure if you're right, but it does seem like most players are posting about what's going to happen in the next year or two. Meanwhile Google announces the release of Gemini Enterprise and "btw, our lead Quantum guy just won the Nobel Prize".

At the point where AI is less of a race (bc a likely winner appears) all the semis/storage/memory trades are burnt toast. That’s why $nvda would panic, gotta keep race competitive at any cost. Not to mention if TPUs played a major role in the win, then $nvda is burnt toast.

Did something happen? In the last 3-4 weeks new things: - memory severe shortage - storage severe shortage - OAI appears in panic mode to get long term strategic investors / compute - $orcl $nvda seem in panic mode to enable oai - massive compute shortage consensus

@RyeNotBerben I always attribute this to you. lol

"if you can't identify the idiot, there's a good chance it's you" --unattributed

"Don't be afraid of a stock because it's working" --Julian Robertson

@nnzp1730 Think they called it the Viking Takeoff. You're right, wasn't uncommon. Don't fly as much anymore so not sure if it still happens.

I was such a F14 fan as a kid (despite going to STL to watch the F15s go vertical on a regular basis)……I love this 👇 account.

@SheDrills Most solar energy production is in the form of electrical current and global energy consumption is around 20% electrical. This may be the nuance in his statement. Could have been worded better.

Trump administration math is embarrassing.