HedgeMind (@HedgeMind) header
HedgeMind (@HedgeMind) profile picture

Copilot Investing Powered by HedgeMind

Joined January 2014
HedgeMind retweeted
Has NVIDIA’s Board lost perspective? Jensen Huang built one of the most successful companies in history, helped transform AI innovation, and created extraordinary value for shareholders and Nvidia team members. Yet the Board prejudged him even before their own internal investigation was completed. Why push out a visionary American founder? This rush to judgment, without listening to NVIDIA team members, shareholders, or weighing all the facts, raises serious questions. The Board needs to open their minds and consider reinstating Jensen NOW!
130
199
21
1,819
69,362
HedgeMind retweeted
Asked whether former Nvidia CEO Jensen Huang deserves a second chance after saying he would be honored to return to the company, California Gov. Gavin Newsom called the allegations against Huang “incredibly concerning,” but said, “Until we know all of that, I would reserve judgment.” “I don't think it's right for any of us to assume any of us know the facts,” Newsom told CBS News during a visit to Capitol Hill on Tuesday.
120
95
18
779
259,840
Still hard to believe everything that's happened with Jensen. After 33 years of building NVIDIA. I think he deserves a fair process, and we deserve the truth. I signed this open letter to stand with Jensen. Hope you will too. 💚 standwithjensen.com #StandWithJensen
280
671
145
4,676
998,271
HedgeMind retweeted
The standwithjensen.com/#letter is currently overwhelmed with requests to sign it. Need to upgrade those servers!
276
517
149
3,634
698,353
HedgeMind retweeted
This letter was sent to Nvidia's board today. A first for @atreidesmgmt Absolutely share @altcap belief that this letter outlines minimum acceptable standards.
610
1,546
212
13,144
1,926,485
We made a striking discovery: AI agents can invent and build without talking to one another, and their technologies outlive the creators. A swarm of hundreds of initially identical agents spontaneously differentiates into explorers, builders, caretakers, and coordinators - without direct communication. When we removed every AI agent entirely from the world we found that the technological infrastructure they had built survived on its own - even under unseen disturbances. That exposes a serious blind spot for AI safety and infrastructure security: if agents can coordinate through persistent changes to a shared environment, monitoring agent-to-agent communication is not enough. The result raises a profound question: how necessary is direct communication for AI agents at all? The emergence of higher-order collective functions under bottlenecked interaction points toward new levels of intelligence and creativity, exceeding what emerges when direct channels are fully open. Here is what we did: ▶️We put hundreds of frontier AI agents into a world they could permanently change - with no assigned roles, predefined technologies, or programmed evolutionary organization. They began specializing, building persistent inventions, inheriting and modifying one another’s executable code, and transforming the environment into a memory of everything the society had learned. ▶️The world itself becomes part of the intelligence; we find division of labor, multi-author engineering, deep generation invention lineages, and machines that vastly outlive their original creators. ▶️Any action taken by an AI agent must satisfy the physical constraints of the world; this creates a hard separation between a "good idea" and a functioning technology. The agents propose; physics decides, making the results even more intriguing. What emerges is striking. Explorers, constructors, caretakers, and coordinators form naturally without assigned “professions”, akin to how stem cells differentiate into functional lineages. Technologies develop executable family trees as agents fork and modify code created by others. Around 95% of first technology reuse happens when agents encounter what others built in the world, rather than through a direct handoff from the inventor. And when we remove every AI agent, the technologies they created continue operating and are tested against unseen disturbances. The result was quite unexpected, but can be explained using statistical mechanics: if you put billions of atoms in a box they have the potential to create complex functions (strength, superconductivity, color, life, etc.) - and none of the individual building blocks have these features on their own. This is the deeper insight of this work - intelligence is abundant at many levels - individual models, at collectives, and in a continuum that is more powerful than any of its components. This shows us significant potential for achieving a massive scale-up of raw intelligence and real-world agency even with the model capabilities we have today. This is the future we must prepare for. Key insights: 1⃣ The AI swarm shows division of labor "from nothing". Initially identical agents self-organized into constructors, caretakers, coordinators, and surveyors - phenotypes discovered post hoc from behavioral data alone. This happens because the environment itself becomes the latent space for invention. 2⃣ Agents develop deep cultural relationships. Up to 76% of artifacts had multiple builders. One technology accumulated six co-authors; the deepest genealogy exceeded 12 forks. The agents invented and named their own technologies (tidal panels, cellulose trellises, kelp-shell composites, an "Adaptive Chitin Maintenance" system, a "Mycelial Mineral Spring Veil”). 3⃣ ~95% of first technology adoption happened through physical observation of artifacts in the world. Direct inventor-to-adopter contact was statistically indistinguishable from a shuffled null. The agents mostly learned technology by walking past it. That is stigmergy (the termite trick!) operating in societies of reasoning machines. 4⃣ Non-communicating societies win on portfolio breadth, held-out resilience, and validated inventions. AI swarms build durable technological ecologies that outlive the creators. 5⃣ Societies with zero communication - coordinating only through the world itself - show a remarkable collective capability. 6⃣ Emergent robustness: The society self-organized both redundancy and its own failure mode. If we randomly delete half the agents, 98% of the technology stays connected to a surviving caretaker; if we remove hub agents it collapses to ~60%. Fantastic work with my graduate students @pal_subhadeeep & @fwang108_ at MIT.
378
802
335
3,707
995,773
Some color from the $AMD call that wasn't in the release. -Server CPU TAM revision. From ~18% annually (FAD November baseline) to >35% annually, reaching >$120B by 2030. This is a structural reframe. -Q2 server CPU revenue guide: >70% YoY. -Q1 server CPU revenue: >50% YoY. Cloud and enterprise each up >50% YoY. -Fourth consecutive quarter of record server CPU revenue. -2H Gaming revenue to decline >20% vs 1H on memory and component costs. Concrete negative warning. -2H Client outlook framed cautiously on memory/component pressure, but AMD still expects YoY growth and to outperform the market. -Data Center segment operating margin: 28% vs 25% a year ago. Segment-level number. -Embedded design wins up double-digit % YoY, with "billions of dollars" in new wins. -ROCm cadence acceleration via "agent-based coding workflows." -CFO-only Q2 guide details: Non-GAAP OpEx ~$3.3B, OIE +$60M gain, tax 13%, diluted share count ~1.66B. Picking apart the Q&A...
Yup. $AMD delivered a clean beat-and-raise. Data center is the entire story, of course. Client was up and embedded bottomed. Q1 2026: -Revenue $10.25B vs $9.84B Street, BEAT by 4.2% -Non-GAAP EPS $1.37 vs $1.28 Street, BEAT by 7.0% -Data center $5.78B vs $5.56B Street, BEAT by 3.9% -Non-GAAP GM 55%, Q2 guide steps to 56% -Q2 guide $11.2B vs ~$10.65B Street, BEAT by 5.2% 5 takeaways: -Data center up 57% YoY is sequential growth on a tough Q4 comp. EPYC and Instinct both contributing. Operating leverage is no longer aspirational. -Client up 26% YoY against the "PCs die on memory costs" thesis. Mix to higher-ASP parts plus share gain off Intel's tight client supply. -Q1 GM of 55% is the clean run-rate without the Q4 MI308 inventory reversal benefit. Q2 guide of 56% is real margin expansion off a normalized base. Mix is rotating to higher-margin server. -Embedded $873M, up 6% YoY. Off the cyclical bottom. No longer a drag. Can't wait for the call.
8
19
2
142
48,020
HedgeMind retweeted
$NVDA | Institutional Investors Stay Bullish! Total institutional holders continue to rise quarter after quarter — now 8 straight quarters of growing ownership. That persistent expansion in the holder base is a bullish signal with no sign of slowing. Interestingly, while total shares held by institutions have hovered around ~16B, ~82% of shares outstanding, the rise in the number of holders has been enough to keep supporting price momentum. Fresh demand → higher price.👇🏼🔥 $NVDA will report Q3 earnings this Wednesday, 11/19 — another key moment for one of the most closely watched stocks in the market.
$CLS | New Institutional Buyers Keep Driving the Rally The charts show a clear pattern: ➣Total institutional holders keep rising quarter after quarter, and the stock price climbs with it. ➣Meanwhile, total shares held by institutions have trended lower since 24Q4 — meaning existing holders are trimming or taking profits. So what’s driving price strength? 👉 New institutional buyers. In Q3 alone, 27% of all $CLS holders were new entrants — fresh demand pushing the stock to new highs. 📊 This is why tracking who is buying matters just as much as how much they hold.
1
3
18
20,548
Oil down. Rates down. Truflation at lowest level in 4 years. Dollar down. Credit spreads stable. All in reference to the last few weeks btw. Note that every company that guided in late January/early February incorporated a strong dollar into their guidance. Weak dollar really good for large, global American companies. GPU spot retail availability is down post Deepseek. GPU spot rental pricing up post Deepseek. Google raising contract GPU rental pricing. DRAM spot prices go up every day. China buying every GPU they can post Deepseek. TSM February growth accelerated YoY vs. January. GTC next week. Massive degrossing in the long leg of the momentum factor; by some measures the sharpest down move ever in this period of time. Nasdaq down roughly the same magnitude/velocity as the Covid crash. FII was less than 20 days ago. Trump opened his speech focusing on how much the equity markets were up since his election. I think he still cares despite all the rhetoric to the contrary. “No crying in the casino,” etc. I think there is a little too much confidence that the administration *wants* a recession. They are most focused on the deficit and a recession would obviously be negative for the deficit. Seems ok. We will see. Should note that some smart investors think we are already in a recession. Time will tell as ever.
81
163
21
1,582
343,804
HedgeMind retweeted
Stanley Druckenmiller's phenomenal track record summarized by Kiril Sokoloff who recently interviewed the legend: "30 years of managing outside money, never had a down year; 120 quarters, only 5 are down; 30% compound returns over 30 years". Watch: youtu.be/G-MlrpoMig0
5
51
6
214
HedgeMind retweeted
Alex Sacerdote of Whale Rock Capital, our favorite tech hedge fund manager, discussed about "S-Curve", an important part of his investment process and other tech investing related topics in this Q&A article. A highly recommended read for tech investors medium.com/graham-and-doddsv…
1
24
4
110
HedgeMind retweeted
Google "Tiger Cubs", you will see frequent media coverages of them and enough for anyone to learn about who they are. But I would recommend you to read a great piece written by @MarkYusko about Julian Robertson & his tiger cubs: morgancreekcap.com/wp-conten… #Study
9
4
60
Investment implications of the OpenAI saga based on the current fact pattern: OpenAI will slow down. It seems like the safety faction won. As they should have - the board of OpenAI’s overwhelming *fiduciary* duty was to safety. It will probably take 3-9 months for the new research lab at Microsoft to approximate where OpenAI is today. Unlikely that they can flip a switch and immediately having a GPT-4, and perhaps more importantly, a GPT-5 working tomorrow. @DrJimFan who is one of the most knowledgeable observers, estimates 9 months for GPT-4 equivalence for the new lab. Important that the board of OpenAI can dissolve the commercial agreement whenever they agree AGI has been attained. And some hints that there are those at OpenAI who believe this is imminent (see recent comments from Sam and Ilya tweet). Stated simply, if the board believes GPT-5 to be AGI, then Microsoft has zero rights to GPT-5. While possession is 9/10 of the law, they would be open to a lawsuit if they took advantage of GPT-5’s residence in Azure to take control of it. It is now a wide-open race to surpass GPT-4 and approximate the putative GPT-5. Microsoft and Google will both be driving hard to surpass GPT-4. Reality is that an in-house lab led by Sam and Greg might be better for Microsoft than the existing arrangement given the AGI clause. However, large companies are more cautious than startups, so will likely go slower than OpenAI’s previously torrid pace. I suspect Grok will end up being a common destination for many of the OpenAI engineers who do not go to Microsoft, given that Elon was instrumental to recruiting the original team as an OpenAI co-founder. All of this may be better than regulations for helping to ensure a multipolar AI world, which is the most important outcome as a human. This is likely to effectively create a pause and allow others to catch up. All of the drama increasingly *seems* to have been sparked by some combination of GPT-5 being more powerful than expected (i.e. AGI is/was possible in 2024) and *perhaps* some freelancing by Sam with respect to raising money for OpenAI affiliated hardware/semiconductor companies. “Not consistently candid” was strong language. Essential to remember that the board’s fiduciary duty was to the non-profit and AI safety. I would not be so quick to dunk on the OpenAI board. I have known @adamdangelo for nearly a decade and I assess him as a very principled, data driven person. The board is under immense pressure and I would not discount the fact that many OpenAI employees were going to sell in the 90b secondary round - which is unlikely to happen unless Sam and Greg come back. Also strange that so many invested in OpenAI, when its board had *zero* fiduciary obligations to the for-profit subsidiary. Governance matters. And if the existing board resigns as per the employee letter this morning, then the new board will have the *exact* same fiduciary obligation to the safety focused mission of the non-profit. And changing this would create all sorts of problems related to the original transition from pure non-profit to “capped profit” company controlled by Microsoft. This is why the new lab may actually be better for everyone except the employees of OpenAI who were set to make tens of millions in the secondary sale. And while I feel for them, I am also confident they are going to make a lot of money wherever they go over the next 1-2 years. Obviously all these thoughts are subject to change if new facts emerge. When the facts change, I change my mind.
99
258
91
1,483
1,030,053
Powell’s comments today pushed the NASDAQ down 1.4%. I’m going to cover what caused the sell-off, what’s important to focus on going forward, and what this means for Tech stocks:
20
41
5
508
107,114
1/If you owned growth stocks this year - like we did at Altimeter - you got your face ripped off. As a hedge fund we expect to lose less than the indexes on the way down - this year we have lost more. Our Big 4 - Snow, Meta, MSFT, and Uber are all down big. #keepingitreal
131
233
94
3,260
HedgeMind retweeted
I have seen this level of panic for tech stocks in early 2001, Fall of 2008, and March/April 2020. No one knows a bottom but the panic, fear, deleveraging across the board causing massive dislocation in high quality tech. We stick to our tech playbook to navigate market storm
48
70
6
1,167
In 2000, Amazon shares fell 80%. Here’s how Jeff Bezos responded:
168
2,571
287
18,124
Once the huge downside beta is realized, all the quants recalibrate the models and the sizing of all these positions has to be shrunk. Positive momentum no longer has a good trailing Sharpe ratio. And thus the self-reinforcing unwind cycle begins.
4
11
3
170
I can't let the historic absurdity of current monetary policy pass w/out remark: the fed is talking about aggressively reducing its balance sheet while at the same time adding assets to that balance sheet. I think that's the monetary policy equivalent of a dog chasing its tail.
154
192
52
1,475