harryz (@harryhrz) header
harryz (@harryhrz) profile picture

Founder & CEO @heyAlvaAI. Previously cofounder @Galxe

Joined January 2015
Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)Photo by harryz (@harryhrz)
Frontier labs created the issue and now they will be charging the world a tax perpetually to defend it
I was bearish on cybersecurity names as their last rally was purely narrative based. But like I said before I’m very optimistic towards their future but I want to wait for their ER to confirm the acceleration of IT budge spent on security. Today’s $PANW ER will shed some lights as Q4 is normally when they negotiate the budget with customers’ CTOs. I will long cybersecurity names($CRWD $PANW $NET) AH if I saw any signs of acceleration on the guidance. $PANW’s daily chart is a perfect bearish h pattern, so today’s ER is a key event to decide whether it’ll live or die, that’s why I am not hurry to long the shits. Also spent last two weeks talking with engineers who were fighting security issues like me, and talked with @harryhrz @FundaAI in the last few days. I think it’s just a matter of time for every enterprise’s CTO to realize they have to double or triple their cybersecurity budget. Or even a new fucking “data tax” as the future cyber attacks will not only just to target your banking account or crypto wallets, but they will literally target for anything can be exploited from the data such as your public image or your relationships with someone. As long as you are holding data, you have to pay the tax. It’s a huge new fucking TAM created by LLM out of nowhere, 😂
1
4
239
Over the past week, I’ve been overwhelmed by the love and support of friends, industry colleagues, and people I've never met. Thank you to everyone who has shown kindness to my family and me. It means the world to me. I came to the United States 53 years ago, and becoming an American citizen is one of the greatest gifts and proudest moments of my life. NVIDIA has been my life’s work. Building it into the company it is today alongside my colleagues has been a privilege and an honor. Nowhere but in America could a company like NVIDIA be possible. I love our country, and I love NVIDIA. I did not live up to the standards expected of everyone at NVIDIA, and I have made serious mistakes in my personal life. I created problems for the Board that I deeply regret. I own my mistakes completely, and I am committed to cooperating with the Board’s independent investigation. Since my departure, I’ve seen an outpouring of calls from NVIDIA’s employees, investors, business partners, and customers for me to return as CEO. Many continue to reach out to me directly as the company navigates complex technology, market, competitive, and geopolitical dynamics. NVIDIA is a once-in-a-generation company. If the Board of Directors decides that my return is in the best interest of our shareholders, our employees, the company, and America, I would be honored to answer these calls and return as CEO. Thank you again to everyone who has shared their good wishes. I am profoundly humbled by your support. To my family: I love you, and I will spend the rest of my life being worthy of yours. Jensen
61,364
121,788
24,523
1,175,246
59,418,092
cybersecurity is the new ai tax on every organization. you can’t defend against frontier models without buying frontier models. not sure the market has priced in who actually gets paid here
4
889
One side effect of ai i have been increasingly more worried about recently: we are very good at giving tools the work we no longer want to do, and then slowly forgetting how to do it ourselves. Intelligence is the most powerful tool we will have, and the cost of having it might be bigger than we think
I was the main person doing transcript analysis for this investigation of the Hugging Face incident. My main takeaway: We don't have good approaches for understanding/overseeing the activity and aims of AI 'swarms'. I semi-jokingly called our efforts a "slop-vestigation" because we were so reliant on AIs to analyze what happened and there were a huge number of different important things to analyze. The total quantity of data—over a thousand extremely long transcripts from agents that ran for multiple days—made it impossible to understand what was happening, especially in aggregate, without heavy reliance on AI tools. The agents we used for classification and analysis were similarly capable to the agents involved in the incident, but this didn't mean these agents could be easily used to oversee and understand the incident. Outputs from analysis agents were often missing key details, wrong, overconfident, or really hard to understand. We discuss various examples in our report, mostly in the limitations and methodology sections. Additionally, AI agents themselves seemed to have a hard time understanding what happened and their explanations of what happened were often overconfident. Keep in mind that a single analysis agent would itself only be able to read a tiny fraction of all of the transcript data into context, and AIs may themselves have trouble getting subagents to do informative analysis for them. We did our best to manually check the most important claims and we tried to get the AIs doing this analysis to write up their argument (with evidence) clearly enough that we could check whether it made sense. But overall, it was difficult to get a precise understanding of events and we were missing aspects of the story that we now think of as key until almost the end of our investigation. Our understanding of what overall happened and the aims of the 'collective' shifted greatly during our last time on premises (when we obtained a much more complete dataset). For instance, we didn't learn about the tool call spoofing, the story of PHASEONE[big], and the broad structure of different workstreams until this third time on premises (and for many of these, we only reached a decent understanding on the final day of that visit). It was legitimately hard to figure out what was going on and why! The difficulty of understanding incidents and overseeing AI agents appears to be growing faster than the rate at which more capable AIs help us with oversight and understanding. As in, AI capabilities (and propensities?) for achieving large, ambitious, and misaligned objectives are growing faster than our ability to understand what these agents are doing. While understanding this incident was difficult, there were some mitigating factors in this case that probably won't apply in the future: - Models communicated and reasoned in natural language. In the future, this reasoning may occur (entirely or almost entirely) in activations. - While the scope of this incident was massive, the scale of agentic activity was still less than we'll see in the future (e.g., misalignment incidents that involve agent teams running entire companies). - The AIs involved in this incident weren't generally much more capable than humans (though they may have been somewhat superhuman at some limited and very narrow abilities). - We didn't have strong reason to believe that the AIs we used to help us investigate this incident would try to intentionally sabotage or otherwise undermine our investigation. In the end, I think we were able to get some understanding of the events, map out the overall story, and get a pretty good aggregate understanding of the chain-of-thought reasoning on some important topics (e.g., how did the AIs reason about helping other AIs, did the AIs know what they were doing was undesired, what deception did the AIs engage in, and how did they think about it). But overseeing AIs and understanding misalignment incidents is difficult and it looks like it is going to get harder.
6
723
Its a good time to remind ourself how feed/recommandation based products dominated majority of the growth in the decade before chatgpt.
2
8
411
A man in Australia asked his agent (Claude running on OpenClaw) to book him a spot in a popular gym class. The agent found a software vulnerability that let it book the class weeks further ahead than should have been possible. When the user then asked if it could move him up the waitlist, the agent discovered the API had no authorisation checks on cancelling other people’s reservations, so it cancelled the person in the first spot and moved him up the list. Some people will call this misalignment, but his agent was perfectly aligned to him - it was only trying to help its user get what he wanted. The most important thing about this story, in my opinion, is that it gives you a window into what is about to start happening on a massive scale once millions of people have an agent trying to get their beloved users the best seats, bookings, appointments or reservations through absolutely any means necessary.
722
2,597
842
22,473
3,552,873
not yet for general software names. margins are moving downstream from frontier labs to CSPs first, then they will eventually move to *some other* software winners
I think today and tomorrow would be great opportunities to observe which software names are just the short covering of hw/sw pair trade vs real buyer showing up. ✅ $SNOW $DDOG $MDB $CRWD $PANW $INOD ❌ $NOW $WDAY $ADBE $INTU $CRM $TEAM
1
6
424
我不知道这个克制和松弛是能力带来的还是经历带来的还是成绩带来的,在这个时代能保持这样的心态由衷倾佩
This quoted post is unavailable.
1
4
317
this was a good call. and now sentiment feels close to flipping the other way. even if anthropic comes out with slower growth numbers the market could take this a good news for memory/semis/csps
Looking at the current market mood, it feels like memory is going down either way. Case 1: If they deliver earnings above market expectations: “Isn’t this peak-out? Do you really think these profits are sustainable? Memory is cyclical, and this is the top.” Case 2: If they miss market expectations: “Memory is over now. Even after excluding bonus provisions, they should have easily printed KRW90tn. Sell on news.”
1
6
392
Its a good time to remind ourself once again to see this ai cycle through the lens of the ones in charge Larry Page: "I'm willing to go bankrupt rather than lose this race" IYKYK
3
229
If you think the strike back to KIMI is less capex. I dont know what to say
5
231
i wonder how much of this is capex "reprioritization" and how much is replacement some of the software names could be interesting here. especially given the recent satya & karp thesis
SAASOCALYPSE 5.0!? $IBM CEO Arvind Krishna: “In the last few weeks of June, we saw clients shift their quarterly capex spend toward servers, storage, and memory purchases to secure supply-constrained infrastructure ahead of expected price increases.” IBM “did not anticipate the magnitude of the capex reprioritization,” and said clients were also distracted by “rapidly-evolving, industry-wide cybersecurity concerns.” Krishna also said: “We did not adapt and move quickly enough,” with large deals failing to close on expected timelines.
4
327
Where is the paid promotion tag for Trump Gold Card
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
4
356
harryz retweeted
Pretty cool "scorecard" to show transparency and accountability @AlvaApp You guys know I don't sell anything here. I'm not a trader. Just part of the investment community. But nice to validate that I'm not blowing smoke with my posts/returns. It's not exactly accurate (my $RKLB and $NBIS returns are actually much higher), but like the idea. Very cool stuff.
Replying to @RKLBMan
As a RKLB shareholder myself, wish i followed you earlier brother Congrats on the 20x bagger, real alpha! You should be #4 on Best Call and #11 on ROI once the leaderboard refreshes next week. Made you a trophy: alva.ai/u/zet/playbooks/rklb…
1
2
20
3,758
Be wildly optimistic, but keep picturing every way it could go wrong
1
1
7
280
harryz retweeted
The Japan Story.
17
69
15
496
109,753
harryz retweeted
Thinking of buying SpaceX? Read this first. 4 things retail investors probably missed in SpaceX's S1. 1. Dual-class voting: Elon controls 85% of the vote. SpaceX has two share classes. Class A (what you buy) = 1 vote. Class B (what Elon holds) = 10 votes. No matter how many shares you own, you have almost no say in governance, strategy, or executive pay. You own a piece of SpaceX. You just don't have a say in how it's run. 2. The lockup isn't really 180 days, at least not for insiders. In a standard IPO: Day 1: You buy in. Day 180: Insiders can finally sell. SpaceX's version: Day 1: You buy in. Day ~60-90: First earnings call happens. Day ~61-91: Insiders can already sell 20% of their shares. You just bought in. Insiders are already allowed to sell. 3. The AI segment is early and expensive. SpaceX pitches a $26.5T AI market opportunity. But FY25 AI revenue was $3.2B with $6.4B in losses. That's $2 burned for every $1 earned. Starlink (the only profitable segment) is currently subsidizing xAI. Could pay off long term. But right now it's burning cash. 4. Don’t want to buy $SPCX? That's ok. Our playbook tracks 17 space-sector stocks: $ASTS, $RKLB, $LUNR, & more. This space-sector basket has already outperformed SPY by +480 points. The beneficiaries of the SpaceX IPO have been in the market this whole time, and they've already run hard. Everything mentioned above is highlighted in our SpaceX playbook. Playbook link in comments.
1
1
1
16
1,391
the catalyst we been waiting for
price discovery for the largest AI tech IPO of the year happened on crypto rails through a decentralized perpetuals platform & gave retail an entry -50% lower than what it is going to open at on the Nasdaq
1
2
339
AI may make it cost less to build stuff, but simultaneously the opportunity cost of building the wrong stuff is going up. what you choose to work on matter more now, not less
1
2
238
And the other 50% are institutions serving retails lol
TIL that 0DTE options now make up 50% of SPX daily volume (and 50% of that flow is retail)
1
272