Menlo Park, California
Joined June 2011
Subquadratic has released SubQ Snap 1.1, tying d1-ultra and Elephant 9B for first on the Artificial Analysis Decision Index v1.0, driven by a low hallucination rate SubQ Snap 1.1 is the new decision model from @subquadratic. It is priced at $0.04 per 1M input tokens and supports a 384K token context window, compared to 32K for d1-ultra and 16K for Elephant 9B. Key takeaways: ➤ SubQ Snap 1.1 scores 65 on the Artificial Analysis Decision Index v1.0, tied with d1-ultra and Elephant 9B: the three models are 8 points ahead of the next model, Jev 1.14 (57). ➤ SubQ Snap 1.1's AA-Omniscience-Decision result is driven by a low hallucination rate: its non-hallucination rate of 86% is second only to Elephant 9B (93%), ahead of d1-ultra (85%). Its accuracy is 20%, compared to 32% for d1-ultra and 12% for Elephant 9B. The non-hallucination rate is the share of questions not answered correctly where the model chose "I don't know". ➤ SubQ Snap 1.1 scores 18% on HLE-MultiChoice: level with Elephant 9B (18%) and ahead of d1-ultra (15%). HLE-MultiChoice is the text-only multiple-choice subset of Humanity's Last Exam. ➤ SubQ Snap 1.1 costs $30 per 1M tasks: comparable to d1-ultra ($32), and ~3.6x the cost of Elephant 9B ($8.4), which scores the same 65.
14
10
5
162
19,029
Michael Olmstead retweeted
Plug and Play has reached 40 unicorns 🦄 With this milestone, we're proud to announce we have over 2% of the world’s unicorns in our portfolio. We reached this accomplishment earlier this summer when Tensordyne crossed a $1 billion valuation, becoming our 40th unicorn. #PNP
2
6
934
Michael Olmstead retweeted
SemiAnalysis numbers imply OpenAI is selling Cerebras-powered Ultrafast inference at ~$200M per megawatt per year. People are re-running the maths hoping for a typo. We're not surprised. A megawatt is a megawatt. What it earns depends on the silicon inside it. Here's how we think about it: Near term, revenue per megawatt wins. Power is the scarcest input in AI, and whoever monetizes each MW best gets the next MW. Long term, gross margin per megawatt wins. Prices compress, competitors arrive, and only the operators with real cost and pricing advantages keep the spread. Cerebras is built for both. Speed is the product. A frontier model at 750 tokens per second is something GPU clouds haven't publicly matched, and customers like Jane Street pay a premium for it. Premium pricing on a contracted, scarce resource is how you protect margin when everyone else is racing to the bottom. That's why we're extremely bullish on Cerebras, and why we're buying as many megawatts as we can. Premium inference unlocks the most delightful user experience in AI. No more waiting, waiting, waiting, staring at little thinking animations while the model works in the background. Answers that keep up with your thoughts. Our customers are even more excited than we are. That's General Compute.
19
36
14
248
68,937
Michael Olmstead retweeted
The pace of change in the silicon world is crazy rn A year ago 3 of the 15 inference asic programs we looked at had shipped silicon in volume. Now that number's 9. Each of them have their special sauce or angle. Some compete with Nvidia - some compliment like the memory focussed folks Mini-Cambrian explosion happening and i'm quite sure most of them will be roaring successes Shipping in volume with failure testing, rack scale solutions is where our lenders, not just a VC, can underwrite the growth at scale This year is going to be huge. Next year even huge-er
4
1
17
903
Michael Olmstead retweeted
🚨 BREAKING: Super excited to announce we're deploying the world's fastest inference with @Cerebras. Talk to any developer and they're excited to build with 20x faster AI... the problem is there's almost no compute available. We're here to solve that. Using GPUs for prefill - it's now more affordable than ever too. Thank you to the whole Cerebras team and excited to grow this partnership. First tokens live Q127 🚀
67
108
62
813
1,059,489
Michael Olmstead retweeted
If you’re in AI, pivot to crypto
483
518
149
6,492
548,911
Michael Olmstead retweeted
bitcoin exchange reserves at a 7-year low. long-term holders sitting on 16.1m BTC, all-time high. corporate treasuries absorbed 175k+ BTC in 2026 alone. and now coinbase institutional says 100+ countries have committed to sovereign BTC strategies. only 2 have disclosed holdings so far. bhutan proved you can mine with stranded hydro at zero cost basis and nobody even knows until arkham finds your wallets. that model scales to every country with surplus energy and weak financial infrastructure. entire nationstates will soon begin buying up your bitcoin and most of them won't announce it
64
17
677
60,871
Michael Olmstead retweeted
The Data Foundation has invested $3 million into Kled, bringing our total funding to now $14 million. We’ve been extremely selective with our investors and have only taken strategic capital. The foundation will be joining some of the top venture funds and angels who have all backed Kled as the world’s leading human data marketplace. Kled is on track to do over 5 billion user uploads and have over 2 million daily active data contributors this year. Onwards and upwards.
66
40
6
498
80,466
Michael Olmstead retweeted
MSTR getting extremely oversold. If you were listening to me since Q4 2024 you never got overly exposed to this. All that said, at these levels the bounce could be explosive. Nothing fundamentally wrong with Bitcoin. Fear always fades. Don’t get sucked in to the mania either way.
160
140
8
2,890
229,285
Michael Olmstead retweeted
Bitcoin will be fine. Stop acting like a pussy.
256
179
29
2,902
124,233
Proud to be involved with so many amazing companies in my nearly decade and a half career! @PlugandPlayTC
35+ unicorns. Every one of them talked about more than the investment. First customers, first U.S. offices, introductions that opened doors they couldn't have knocked on alone. youtube.com/watch?v=O1sQOyJ5…
1
3
335
Michael Olmstead retweeted
Congratulations @ElonMusk and $SPCX on a historic IPO. Thanks to you, 25% of the Mag8 now holds Bitcoin on the balance sheet.
991
2,183
150
29,800
1,174,174
Michael Olmstead retweeted
Puzzle pieces are starting to fall into place. Gradually, then suddenly. $BTC
66
119
18
2,281
113,912
Michael Olmstead retweeted
the last 3 years I've played in a big member-guest golf tournament at my old home course almost everyone in the field has done pretty well with their business or investing so when crypto would come up there's ALWAYS someone who would ask: >is it too late to buy Bitcoin? >do you like XRP?? >what coin should I buy? but not anymore not a single person in any of the 6 matches and I probably talked to 50+ other people throughout the weekend no comments about prices being down or that memecoin they lost money on or whether the token they've been holding will recover no one was still bullish or giga bearish...just not interested at all if the topic came up we have officially reached peak apathy stage
23
4
3
100
23,358
Michael Olmstead retweeted
"What if the model companies do this?" is the new "What if Google does this?" I.e. the meaningless question investors ask that shows either that they're stupid or that they dislike you and are looking for ways to find fault.
189
182
63
2,934
320,359
Michael Olmstead retweeted
Bitcoin 🚀
153
63
6
1,483
115,572
Michael Olmstead retweeted
Most people are willing to buy the S&P 500 above the 200 day moving average But they're not willing to buy Bitcoin close to the 200 week moving average This is why the majority of investors don't make money
47
19
4
322
34,737
Michael Olmstead retweeted
Treasury Secretary Scott Bessent: The Strategic Bitcoin Reserve is "advancing with speed." The Clarity Act will pass "this summer." This is the Treasury Secretary of the United States. Not a podcaster. Not a pleb. The Treasury Secretary.
6
40
7
273
15,985
Michael Olmstead retweeted
Bitcoin's 20% pullback the last week, just did exactly what it was designed to do. Flushed the last of the "cycle is dead" crowd right back into the "4 year cycle is undefeated" crowd. Flip-flop. Every. Single. Time. But here's what nobody is saying — this wasn't even a bad capitulation. It was orderly. Systematic. Almost boring. Exactly like the upside has been. The bull market this cycle has been the mildest on record. So why would anyone expect a wicked 70-80%+ drawdown and an 18-month+ winter? Muted bull markets produce muted bear markets. Risk is symmetric. Most people can only see it pointing down, especially when price is moving against them. Bitcoin is down 50%+ over the last 9 months. The people expecting last cycle's carnage are copy-pasting old trauma onto a completely different setup. Same mistake the AI bubble callers made — took last cycle's valuations, applied them blindly to forward expectations, and called it a thesis. Nothing repeats exactly. Ever. Is the pullback done? Maybe. Maybe not. But we are obviously much closer to the end of this trend than the beginning. The math isn't complicated, it's just hard to see through the pain. Most investors can only find conviction after it's already too late. They just got flushed out again, right on cue. The contrarian who stays optimistically bullish while everyone else leaves — that's where the real money has always been made. It's supposed to feel wrong right now. That's the whole point.
10
6
1
134
14,294