English
Back
Download
Need Help?
Log in to access Online Inquiry
Back to the Top
田_田的投研日记
wrote a post · Jun 9 10:29

Three years of AI-focused investment research

$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$ Gains: When I left ByteDance, I sold some options and converted the proceeds into the following four positions—accidentally ending up with half my portfolio in AI: – AMD (+320%): Started selling half of my calls. Locked in until it reaches a 5x return. – Alibaba (+290%): Withdrew my initial principal when it doubled; holding the rest for the long term. – NVIDIA (+110%): Withdrew my initial principal when it doubled. – AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, making me question whether there’s a flaw in my logic. If it can’t hold its ground, I should reallocate part of it into ETFs like TOPT that won’t get diluted by AI-related index rebalancing. (As I’ve always said: AGIX is the baseline asset of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research.) Losses: – Spent ages researching storage and compute infrastructure but completely missed the biggest storage play. At the time, I had no way to buy SK Hynix, and Micron was still seen as a consumer-focused company with declining core business. There wasn’t much I could do—given another chance, I’d probably still miss it. - Sold a very small amount of EWT due to the US-Iran conflict...
$Advanced Micro Devices (AMD.US)$$Astera Labs (ALAB.US)$$Nova (NVMI.US)$$Kraneshares Public Private AI And Technology ETF (AGIX.US)$
Gains:
When I left ByteDance, I sold some options and converted them into the following four positions, which have now grown—almost against my will—into a half-AI portfolio:
- AMD (+320%): Started selling half of my call options. Won’t sell any more until it’s up 5x from here—I’m holding tight.
- ALAB (+290%): Took out my initial principal when it doubled; the rest is locked in for good.
- NVMI (+110%): Took out my initial principal when it doubled.
- AGIX (+60%): Haven’t touched it. Its weighting hasn’t been adjusted upward despite rising valuations of Anthropic and SpaceX, which makes me question whether my logic is flawed. If it can’t hold its ground, I should probably reallocate part of it into AI-focused ETFs like TOPT that won’t get diluted by non-AI holdings. (As always, AGIX is the baseline benchmark of this era—if you can’t beat AGIX, don’t claim you’re doing AI investment research...)

Losses:
- Spent ages researching storage, computing, and networking but completely missed the biggest storage play. At the time, there was no way for me to buy SK Hynix, and Micron was still seen as a consumer-focused company with steadily declining core operations. Not much I could’ve done—odds are I’d miss it again if given another shot.
- Sold a tiny amount of puts on EWT and RMBS during the US-Iran conflict—ran out of cash to build new positions 🥲
For the logic chain, refer to previously published articles
- The Big Bet: Capex Is All You Need
- Investing in AGI: When the water rises, the fish grow big—targeting a 100x return over 20 years

Prompt sent to Gemini, ChatGPT, and Perplexity:
List the AI-related hypotheses I’ve researched, along with relatively complete timelines (by month) and a summary of each hypothesis, plus core conclusions

All results are fed into my own pipeline:
Organized by broad topic categories, each on a separate page
Each page includes
- General evolution of understanding
- Timeline (month-specific) – Research hypothesis – Core conclusion
- Review of gains and losses
Help me create an HTML suitable for posting on Xiaohongshu, about 15 pages long
Disclaimer: Community is offered by Moomoo Technologies Inc. and is for educational purposes only.Read more
Thumbs Up
3
63K Views
Report
Comments (2)
Write a Comment...
2
3