So I revisited an interesting AI post written by “wait but why”, back in 2015 (10 years ago).
I saw an interesting notion about how progress happens in S-curve – a new technology emerge, slow growth, then mass adoption, then matures.
From my lifetime, I can see how that played out in the internet boom.
And I think now the AI boom is unfolding in front of me. So I want to make some predictions here and see how right I am 5-10 years from now.
The 3 phases of the S-curve
This is copied from the post:
1. Slow growth (the early phase of exponential growth)
2. Rapid growth (the late, explosive phase of exponential growth)
3. A leveling off as the particular paradigm matures
And being in the middle of the S-curve can mislead someone into thinking progress is slow, being in step 1 or maybe step 3.
I think right now in the year 2025, it’s the end of step 3 for the internet boom and step 1 for the AI boom.
And in the past decade innovation does feel kind of slow.
Some people commented the best invention of the decade was tiktok and short reels – which is kind of sad.
Standing from here it’s easy to think that technological innovation would happen at the same pace the next 10-20 years as in the past 10-20.
Timeline of the Internet boom
So the internet boom started around the 90s.
For maybe around 10 years it’s this novel thing and many people are excited.
So excited that it created an internet bubble in the stock market that subsequently popped in 2000.
However, when it popped, it’s actually still in stage 1, where the average joe isn’t using the internet or at least not much.
Thinking back when my mother first started using email, it was already 2002 and she signed up because I was moving to the U.S. and she needed a way to communicate.
The true internet adoption happened after the bubble popped, and an average person started using gmail, doing google searches … then by 2010s scrolling facebook, watching youtube, etc.
It was during 2000-2020, companies like apple, google, facebook grew by 1,000%-10,000%+.
And our lives really changed during that period – the iphone, search engines, social media. from crappy car GPS to google maps. from expensive long distance phone calls to cheap skype calls. From expensive TV and media ads to cheap facebook ads. A smartphone on your hand can access any information on the internet.
Then come the 2010-2020s. The mature stage where even grandmas and grandpas are using smartphones, social media, emails, and watching youtube.
Practically no more huge leaps of innovation can happen on the internet. It’s still growing but nothing is “life-changing” per se.
Iphones went from 4 to 6 to 10 to 14. Video content went from long to short to even shorter.
So standing here in 2025, it seems like nothing has really happened in the past 10-15 years.
The slow growth of the next boom – AI stage 1
The AI boom probably started somewhere around the early 2020s.
Starting with this novel chatgpt thing, which people talked about a lot and I have recently started using it a lot, super helpful.
But the average joe is still definitely not using chatgpt. Nor any significant amount of AI like those cute image and video generation thing.
These AI tools now seems novel, and kind of like cute toys which some people dismiss as “useless”.
For maybe 2-3 years now nothing much seemed like have happened – beside the stock market having one of the biggest bull runs ever.
Progress feels slow and many people dismiss the potential of AI.
And at this rate, it might feel like all the hopes and dreams of AI might not even come true.
Which might cause the inevitable(?) stage 1.5 – the crash where people excitement waned and capital start to dry up.
Also there would be a lot of wasted investment in stage 1 since it’s a lot of R&D. Some unfruitful capex that goes down the drain.
And frankly it is possible that it never reaches stage 2 and it’s all just a fad.
But it is my believe that the AI story is not a fad, and stage 2 is coming soon …
The explosive growth – AI stage 2
My prediction is that the explosive AI growth will start happening in the next few years.
So maybe starting late 2020s into the 2030s.
Autonomous cars will replace every driver. The younger generation don’t need to get driver’s license anymore.
AI agents will serve anyone and everyone including the grandmas and average joes. Any average person’s chores and repetitive tasks will be cut in half or more.
Something like AGI will probably emerge and millions of jobs will become irrelevant. AI can teach, AI can use computers and write code, AI can build and construct.
It will be impossible to separate lives from AI by that time, just like it would be impossible to separate lives from the internet at this point.
And as an investor who invested big into this theme, massive wealth will be created and the rewards will be handsome.
It takes courage and patience but if I have conviction about it, I am going to go deep into it.
My investment strategy
So first of all, as I mentioned in my Feb post, I am increasing my US equity allocation, by quite a bit.
Also I am shifting to a more aggressive strategy within the allocation, and getting rid of “defensive” stuff like BRK, COST, and also trimming down QQQ for individual companies.
I will continue to trim down QQQ to a small size and keep GRNY in there but both for a smaller percentage. (previously maybe 20-30%, probably will trim down to 10-15% total)
Then I will maintain the mindset of “staying invested” and not being afraid of losses and waste.
It will be impossible to time the bubble so I will take the hit with the market, and also buying my way down during the bubble while maintaining good amount of liquidity – hence the large position in gold to hedge it. BTC will be somewhat of a diversification as well (or stocks is diversification for bitcoin??)
I will also try to stay with the bigger players who have at least decent to high potential but also not too risky. For example GOOGL and META are both insane cash machines that are also on the front line of AI development. Having PEs in the high 20s or low 30s, the downside is pretty limited but there is still huge upsides.
AMZN and MSFT are probably good as well although I haven’t looked too much into them so am not invested yet. I might start digging into AMZN later and rotate the QQQ in there.
TSLA is the wild one with 100x+ PE but I am still holding onto it at this valuation. I think the multiples will have to contract eventually though and my guess is, it should probably float as a 40-80PE company with its growth potential.
It might be OK to dip my toes into some not-profitable companies with a great story as well but I will keep that at a minimal, to defend myself in case of the big crash.
I will continue to stay educated and informed about the industry and environment as well. If I notice non-winners or story changes in my portfolio, I will (at least try to) stay rational and cut losses on them. If I notice new stories and potential new winners, I will start small and try to rotate the capital in there.
Looking back 10-20 years later, some will be good investments, some will not be. But the few great ones will generate 10x, 100x return and create generational wealth for me.
Don’t miss out because it can’t be seen
Most people will miss out on it too, because they are too focused on their own experience of the present and be quick to dismiss future potentials.
I like the analogy “wait but why” used:
“If I tell you, later in this post, that you may live to be 150, or 250, or not die at all, your instinct will be, “That’s stupid—if there’s one thing I know from history, it’s that everybody dies.” And yes, no one in the past has not died. But no one flew airplanes before airplanes were invented either.”
At times, especially when the market is crashing, it will feel hopeless and I’d want to jump out too.
Almost everyone will become naysayers then and the companies, even the mega-caps, will become undervalued.
So I am using this post to anchor the future me – that this is highly probably real, and the crash is THE OPPORTUNITY to buy more and become even richer when stage 2 happens.