How I Missed a 200x Gain in Nvidia Stock as a PhD in AI

5/5 - (2 votes)

Between 2014 and 2018, I was working as a PhD student in distributed AI.

Specifically, we were using GPUs to parallelize the training of machine learning models.

Both the penetration of machine learning into every single area of computer science, as well as the rapid growth of using GPUs for AI was very obvious for me and all my colleagues.

  • Data volumes were growing exponentially (still do)
  • Model sizes were growing exponentially (still do)
  • Computation demand was growing exponentially (still does)

In hindsight, my research group and I were seeing OBVIOUS truths that were not obvious to mainstream. The perfect investment opportunity!

You might think we all used our knowledge to get rich investing in NVDA, right? Because we were already talking about investments while being very aware that NVidia was winning in AI chip design.

At the time, NVidia stock was priced at $4, now it’s roughly $877 per stock. That’s a 200x in 10 years! If any of us invested a year’s worth of savings into this high-conviction bet, we’d be millionaires from a single investment decision!

Well, I missed it.

I should’ve held NVDA for a decade and done nothing πŸ˜ŽπŸ–οΈ.

  • +50.43% My Personal Portfolio
  • +10.56% Stocks (S&P 500)
  • +68.92% Nvidia (NVDA)
    • $10k grew into $1.9M in 10y

My colleagues missed it too. Most still work 9-5 “for the man” never quite accomplished their dream of becoming financially free.

In today’s video, I’ve analyzed the flawed logic of the missed opportunity because multiple gems can be learned from my mistake:

How I Missed NVDA Stock in 2014 (as a PhD in AI)

Try our CAGR calculator here:

The efficient market hypothesis (=all companies are fairly valued) and my focus on traditional valuation frameworks like P/E based valuation made me reject the idea of actually taking an NVDA position:

Today, nothing has changed for NVidia:

Check out my article on the AI Scaling laws to learn how we’ll keep demanding more GPU computation:

πŸ‘‰ AI Scaling Laws – A Short Primer

NVidia’s motivation of “accelerated computing” is simple: to keep Moore’s Law alive:

Similarly, the AI chip market will keep growing:

Even within verticals, like generative AI text-to-text generation, we’ll scale up computational effort. But we haven’t even scratched the surface for computationally demanding AI tasks such as text-to-video:

A billion humanoid robots will cause EXTREME computational demand in the near future:

πŸ‘‰ Tesla Bot Optimus: Is $5,000 per Share (TSLA) Realistic?

Finally, the ultimate AI application is self-driving cars. Imagine the compute demand here for AI chips!

πŸ‘‰ [BREAKING] Tesla Car Drives Elon Musk in Epic Livestream (FSD V12): NOTHING BUT NETS–All The Way Down!

If you want to stay updated with tech disruptions and valuation models, check out my newsletter (160k subs!). If you desire to master AI, e.g., to create your own startup, feel free to check out our AI courses on the Finxter Academy. If you’re like many other of our students, this decision to join can be the difference between making $50k per year and $200k per year!