crowdturtle 🐢 (@crowdturtle) header
crowdturtle 🐢 (@crowdturtle) profile picture

Data scientist in big tech. Investor in cash-flowing real estate, special situations in public securities, and blockchain.

NYC
Joined September 2011
Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)Photo by crowdturtle 🐢 (@crowdturtle)
crowdturtle 🐢 retweeted
Permabulls comparing this pullback to the Tariff bluff and Iran war pullbacks is exactly why this market will keep going lower. Not being able to identify the end of a liquidity cycle destroyed many people in 2008 and it's going to do the same thing again here.
Retail investor: I sold everything Friday. I’m telling the whole family at Thanksgiving to get out Me: What’s the plan to get back in? Retail investor: when it calms down Me: What does calm look like? Retail investor: idk... when it stops dropping. when the news is better Me: So when prices are higher Retail investor: ... Me: Did you sell in April last year too? During the tariff thing? Retail investor: yeah Me: The S&P went up 57% from that low. When did you get back in? Retail investor: like 3 months later. way higher Me: So you sold low & bought high. & you’re about to do it again Retail investor: so what am I supposed to do, just sit here & watch? Me: Own great companies at good prices. Keep your options 1-2 years out. Size it so a week like this doesn’t scare you out... Please do not be like most retail investors that get smoked by the market in the long run...
51
26
2
580
32,205
crowdturtle 🐢 retweeted
The Nasdaq's biggest crashes in history: 2000: -78%, took 15 years to recover
2008: -56%, took ~3.5 years to recover
2020: -29%, took 3.5 months to recover
2022: -36%, took 2 years to recover
2025: -23%, took 4 months to recover
2026: -20%... so far @grok this time, how long do you think it will take?
9
1
1
31
14,618
crowdturtle 🐢 retweeted
Always use your unfair advantages. Rich parents, good looks, natural drive, charisma, athleticism, whatever you were handed. Nobody gives out medals for playing on hard mode when an easier route was sitting right there. Take the head start. Don’t make the game harder than it has to be. Use everything you’ve got, then keep adding to it.
12
486
11
3,779
78,607
crowdturtle 🐢 retweeted
$QQQ, last 10 years: 2017 🟢 +31.5% 2018 🔴 −1.0% 2019 🟢 +37.8% 2020 🟢 +47.6% 2021 🟢 +26.8% 2022 🔴 −33.1% 2023 🟢 +53.8% 2024 🟢 +24.8% 2025 🟢 +20.2% 2026 🔴 −1.0% (YTD) Every red year was followed by a big green one: 2018 🔴 → 2019 🟢 +37.8% 2022 🔴 → 2023 🟢 +53.8% Is 2027 next?
20
8
226
44,024
crowdturtle 🐢 retweeted
Consciousness is either an illusion or ubiquitous on a gradient. Difficult to argue that carbon is conscious, but silicon is not.
3,587
1,751
643
18,180
1,221,447
crowdturtle 🐢 retweeted
207
1,322
174
11,860
733,109
crowdturtle 🐢 retweeted
THE MOMENT YOU'VE ALL BEEN WAITING FOR I went day-by-day by hand for both the 2000 and 2007 Bear Markets (from peak to trough) and figured out every single run up and run down, in order, with every single trading day accounted for (nothing was omitted). [data: stooq]
74
414
66
1,396
crowdturtle 🐢 retweeted
“The first step to becoming strong is to kill every fear within you.” - Elon Musk
73
526
27
4,066
74,886
crowdturtle 🐢 retweeted
Jakub Pachocki, Chief Scientist at OpenAI, on the impending capabilities acceleration in 2027: 'We have recently witnessed unprecedented acceleration in another domain. For more than a year, we have been posing open problems in mathematics to our models as one of the ways of evaluating them. Once we upgraded to a model of log-perplexity ~0.76, it started producing groundbreaking results. For many of us that have previously worked in mathematics and adjacent fields, seeing the resolutions to questions we've long thought about coming out of the model one after another in a single Slack thread felt deeply surreal. Having experienced that allows us to extrapolate - such a sudden transformation is plausible in deep learning research too. While this field relies more on experimentation, there is a long history of simple insights being missed for years; and so as soon as 2027 we could conceivably see a similar avalanche of fundamental pretraining advancements.'
We decided to provide more transparency into pretraining progress at OpenAI for the sake of moving toward shared safety bars. In “Perplexity and Intelligence” I argue for common limits based on perplexity and other measurements, with progress beyond them gated on our ability to understand AI behavior. openai.com/index/perplexity-…
76
102
11
1,193
53,469
crowdturtle 🐢 retweeted
AI spend still flat since June. Token prices still in freefall. As predicted in October, the market is starting to panic that the $4T of AI capex is a 0.
Heading into a seasonally weak period for token consumption and spend through Jan with token prices still in freefall. The Jan 26 ramp was agents. The April spend ramp was a change to token based billing which pushed people down the cost curve and sent token prices into freefall. What is left to push revenues up in a hyper competitive AI market with a giant and worsening compute glut? If spend is flat to down through Jan (6 months from the July peak), where do you think AI names trading at 20-50x sales will trade? We are now 3 months past peak AI spend. After 6 months the hopium will have worn off and markets will need to price the sober reality of $4T of AI capex being a big fat ZERO.
27
30
8
249
58,420
crowdturtle 🐢 retweeted
Shared this with my substack chat on Friday… A good reminder heading into the short week for Thanksgiving with light volume.
13
48
10
541
49,669
crowdturtle 🐢 retweeted
David Friedberg says Google appears to have hit the RSI runaway loop, and Anthropic is likely there too. a lot is being held back over safety concerns. The models are probably advancing faster than human evaluation can keep up with.
2
29
2,619
crowdturtle 🐢 retweeted
Simulation theory unbeaten.
While at OpenAI, Dario suggested that immersive videogames might be a solution for the political strife caused by AI-induced joblessness. One of many new details in our story on the siblings leading Anthropic. W/ @berber_jin1 and @keachhagey wsj.com/tech/ai/anthropic-da…
67
65
14
1,007
45,947
crowdturtle 🐢 retweeted
Coming to a software shitco near you
Catastrophic cyber attack/data breach against Japanese software company Plaid $4165, stock -57% in 5 sessions. The service is offline, company says a "fundamental rebuild" is necessary and no resumption until mid-January 2027 (!!). No revenue until then.
3
6
119
10,316
crowdturtle 🐢 retweeted
In 1987 Nasdaq went from +30% to -5.3% YTD In 2000 Nasdaq went from +24% to -39.3% YTD In 2018 Nasdaq went from +17.5% to -3.9% YTD In 2026 Nasdaq went from +18.7% to -2% Wild year
5
13
3
165
22,350
crowdturtle 🐢 retweeted
One of my observations during Covid as well. All of the demand from stay at home and an e-commerce boom, but 2 years later AMZN, SHOP, ETSY all down near or below precovid levels. I misread this cyclicality at that time. This is even more pronounced with commodity producers. I’m not sure these booms are as beneficial to their businesses long term as people might think. Take a look at any internet supplier during 1996-2005. Not just their stock chart but their underlying GAAP earnings (which includes the massive write-downs which are real value destroyed). Boom was first 5 years bust was last 5. Same exercise with investment banks and home builders between 2003-2013. Biggest housing boom in history and more value was destroyed than created in both those groups. Same with shale boom: check oil company fundamentals between 2010-2019. The overinvestment often leads to a digestion phase that often more than offsets the value accrued during the boom. Can’t prevent these cycles because that’s capitalism, and it is hard to believe when stocks like DELL and MU are still up 150% YTD, but I think some of these picks and shovels stocks might actually end up doing worse vs the alternate universe where the boom/bust didn’t happen
Flashback to June 14, 2024. At the time, frontier AI models have zero fluid intelligence. They score around 5% on ARC 1. They can't reliably add two numbers. They make laughable logic errors every other prompt. Reasoning models don't exist. Coding agents don't exist. Claude Code doesn't exist. "Code gen" is a complete joke. Scaling up next-token prediction based on web text data is hitting a wall. What if a time traveler told you that, by November 2026... • AI models would have human-like fluid intelligence, thanks to a new paradigm: test-time induction (modern reasoning models). • Not only would ARC 1 be saturated, but unreleased benchmarks 10x harder (ARC 2) and 100x harder (ARC 3) would be saturated too. • Nearly all code would be written by AI, and models could one-shot complex codebases flawlessly. • AI models would be solving multiple Millennium Prize problems. • OpenAI's valuation would be 15x higher, and Anthropic's would be 85x higher. And on top of all that... • Nvidia would still hold ~80% of the AI compute market. • Nvidia would have grown its revenue and earnings 4.5x. • Nvidia would still be growing 75% a year. • Nvidia would have even higher operating margins (67% vs. 65% back then). Based on that information, straight from the future, would you have bought NVDA? If you answered yes -- congrats. Today, 2.5 years later, you'd be down 5% on your investment (even more after inflation).
8
7
3
126
13,934
crowdturtle 🐢 retweeted
Replying to @fchollet @garrytan
Getting this amount of compute turned on, cooled and working reliably is incredibly hard. The cash shortage is also going to be nuts next year. Even cumulative industry delays of a few months between paying for compute and getting it reliable enough to generate revenue will mean hundreds of billions of incremental dollars just for working capital. Where will that come from?
1,686
1,645
975
19,559
4,047,083
crowdturtle 🐢 retweeted
Elon Musk was right when he said, "Go Get your fucking confidence back. Without it, you will fail miserably no matter how skilled or smart you are."
15
277
6
2,756
77,205
crowdturtle 🐢 retweeted
Getting closer
What about the parallel universe where AI is actually just deflationary and highly competitive and no one makes any money off of it?
18
8
232
19,544
crowdturtle 🐢 retweeted
It is amazing that human language turned out to be so unreasonably effective at modeling the universe that training a system to predict what comes next in a sentence would get us this far.
529
1,104
194
17,137
485,859