I worry that the old advice of just investing in index funds may be obsolete as tech companies lose their advantage.

My thesis is that the moat of large tech companies will be removed as everyone gets access to powerful large language models. Everyone will have the ability to become a FANG engineer. I feel like their margins will drop as anyone can create custom software.

FB and Alphabet killed local newspapers and advertising. Uber killed local taxi services. They were extractive businesses as silicon valley earned a tribute for providing software. Now anyone will do it. Local businesses will easily be able to create custom solutions. Business will move away from large corporations towards indie developers and small businesses.

For decades the value in indices has moved from tangible to intangible business. I feel like this will reverse. The business of "doing" will be more valuable.
Perhaps Amazon, Alphabet and Microsoft will end up like IBM of old, and earn their money from hardware via server space on their cloud rather than their software.

I personally use google a lot less now that I have access to ChatGPT. I also program my own solutions now. This will only accelerate. Why would anyone buy off the shelf software anymore?

Maybe we should dump the likes of Salesforce and invest in energy, transport, and companies with physical, capital or regulatory moats. Small businesses will dominate tech, and big businesses will only be for capital intensive areas.

It will flip, the future will be to “Sell atoms, not electrons.”

The whole point of an index fund is that you don't have to worry about this.

The top components of the S and P 500 are completely different than what they were in 1970.

The top 10 components of the S&P 500 in 1970 were:

1 General Motors
2 Exxon Mobil Energy
3 Ford Motor
4 General Electric
5 IBM Technology
6 Chrysler Industrials
7 Mobil Energy
8 Texaco Energy
9 ITT Diversified
10 Gulf Oil Energy

If the income generated by the top companies moves to small enterprises then rebalancing won't fix the issue. Public markets will cease to capture so much revenue.

I wouldn't worry about it.

Dominos pizza and Starbucks revenues grow every year despite other people opening up different pizza and coffee shops.

What are you programming in general? I'm a proponent of the e/acc movement. The big tech companies might try regulatory capture " for safety" and drown out open source LLM's. Biden just issued an order to regulate AI and maybe even graphics cards. I think the big tech companies can always find average joe customers for software models. I mean people have been buying Windows for decades when smaller open source free alternatives are available. Its all about marketing and what phone you use. i-phones aren't going away anytime soon. Big tech isn't going away is my thought but i am hearing smaller open source models are likely the future. That's a low probability best case scenario based on history. What's to stop big tech from buying all the small tech companies? Been happening for decades. Microsoft could even just buy OpenAI. they already sunk their claws in

FYI the best search LLM is Perplexity.AI

I want your scenario to be true actually. Just thinking critically. I'm pretty caught up on whats going on in tech (I think) its moving fast now. Im hearing personal AI agents are the future actually more than anything. I dont see why Apple cant do that. They are the device company. I typed that all out feeling sick as *hit(have slight pnemonia) hope it made sense

I don't see how you are getting from a to b.

Where is the value in building your own FB or Uber? Even if it is completely costless to create your own personal Uber app what is the market? I'd certainly download an app to be driven around by you specifically (I am sure you are a nice guy and excellent driver), but that doesn't seem like even a niche market.

Same goes for FB and Alphabet and other services.

Tech innovations/revolutions have pretty much always moved us towards larger and larger companies. Although not impossible, I'm not seeing any indication this trend will reverse with this new innovation.

There will always be a need for customized stuff, but that need has always been small.


Basically because most needs are best met by mass-market stuff.

It is extremely easy and nearly free to grow your own tomatoes. They are even better in general and you can just as easily be choosy in variety to fit your individual taste. Yet supermarket tomatoes still are a thing.

Clearly developing your own Salesforce software is at least slightly more costly than the existing software and is probably no worse quality ratio than supermarket tomato to home-grown tomato.

(I've no idea whether cap weighted index investing will be better or worse than investing in heavy industrials over the next x years)

If you really think tech doesn’t have moats you probably extremely underestimate their complexity. If you think they’re going to lose market share to individuals creating their own solutions for simplistic issues you probably fail to realize their income is almost entirely based off business solutions. If you think businesses are able to create complex solutions similar to those provided by existing tech companies for themselves you are delusional and never been part of implementing technological solutions at a medium to large business. These companies can’t implement the simplest solutions without using outside services because of the bureaucracy and the issue that those making the decisions are clueless about how technological systems actually work and how to integrate them.

Have you ever been part of implementing or integrating a large scale solution in tech? If anything it’s becoming more monopolistic because cloud/azure solutions are almost required and it’s way more difficult to implement your own solutions without your own servers which is less practical nowadays so people are forced to use prebuilt tech solutions.

i don't really follow the OP so still not sure if i agree/disagree. agree many FANG engineers will lose their jobs to LLMs or like models replacing them. disagree this makes all major tech companies irrelevant or destined for busto. if anything, just reduces labor costs.

google is not resting on its traditional search & ads model indefinitely. everyone's racing to the become the go-to magic AI vendor (which, contra e/accs like Jupiter, imo will lead to peril). i could see scenario where one/few megacorp(s) capture lions share of profits, but that's why you own an index fund. i guess a private company could emerge as the winner and reap all the profits.

Tech has been long-term deflationary as LLM should accelerate that. It's not about everyone creating their own Uber app. It's that the barriers of entry will be so low that every City will have its own competitors. Network effects will determine how that plays out but whatever happens margins will be eroded.

That should play out across every industry. Companies won't be paying $10 for each of their thousands of users for slack because there'll be so many alternatives, and even creating a bespoke solution will be cheaper in the long run.


"Hey ChatGPT look at how we're using Salesforce and create a bespoke CRM for me."


Party because they had so much financial firepower that they could attract all of the top talent and then bolt on any small start-ups. This will be attacked on both sides as their margins get reduced, and the sheer number of competitors should grow exponentially.

I don't think there's a long-term business in being an AI vendor, it's too competitive and moves too quickly. No one is ever going to have the kind of dominance that Google search has had.

Certainly. And that has always caused consolidation, not expansion of the number of companies in that industry over time.

Margins always get eroded over time in an industry. That has always caused consolidation, not the expansion of the number of companies in that industry over time.

I agree that there will be even more apps out there with exactly zero users in the future. That hardly counts as a multitude of competitors.

There already are free alternatives. It is harder to be cheaper than free.

The reason companies use slack is because no one in a legitimate business actually cares about the $10/employee price

Salesforce is effectively a rounding error as far as how much it costs.

The cost of switching to an entirely free crm is far from a rounding cost. Just the hours management will lose listening to sales people whine about changing systems is going to be quite costly.

Except the last part has never occurred that way in the past. This is far from the first time in history that we've had a new innovation.

Companies use Salesforce because they can customize it and immediately apply sales process changes and host the structure and metadata in Salesforce's cloud. We're a very long way out from a LLM creating a customizable CRM like Salesforce at the drop of a prompt.

There's been a proliferation of single developers companies in the AI space. They've been initially focused on new frontiers. But the more people get involved they will inevitably chip away at incumbent industry.

Why would the industry consolidate when the barriers for entry and need for capital drops to zero.


Feels like another Innovator's Dilemma. Only this time it's going to be single devs and small business that will take over the market.

Not one inevitably, but competition exists and today's winners are frequently not tomorrow's. That is absolutely nothing new.

And tomorrow's winners are just as few.



The barriers to entry and capital requirements haven't been a lack of coding ability or the cost of coders for quite a while.

Why, given that coding isn't even remotely the difficult/expensive part, would industries stop consolidating?



If it happens, it will be small companies that become big companies.

Big fish eat small fish. An occasional little fish becomes a big fish. That new big fish eats little fish. If it becomes even bigger it eats big fish.

I guess we're a bit closer


https://finance.yahoo.com/news/sales...201103268.html

And how much of Salesforce's earnings miss was attributable to LLMs?

Having an LLM code for you wont make you a good engineer. It can inspire creativity in good engineers, but bad engineers aren't going to understand the output enough to apply whats good or relevant.

Seems more people are coming around to this idea





https://x.com/aribk24/status/1834402184253038636

he has no idea what he is talking around