I think it’s similar in that…

- Everyone knew the Internet (and now AI) was going to be transformational.

- It was clear it would reshape economies and open up entirely new possibilities.

- But no one really knew what the killer use case would be, or how big any given Internet (now AI) business might get.

So, like a gold rush, huge amounts of money poured in, and everyone hoped their idea would turn into the next massive market.

——

There’s also parallels between NVIDIA and Cisco.

When the Internet was in its infancy, Cisco stock soured because the thinking was that Cisco networking is what’s ultimately going to power the Internet. (Much like how NVIDIA powers AI).

But what happened was that, new networking companies entered the market and it was found that the Internet economics was so large it overshadowed the relatively small networking needed for it in comparison.

Cisco still exists and was a winner for a long time. The company that emerged hugely successful out of the dot com era was Amazon, they dominate ecommerce today.

Take a look further back to the Personal Computer era in the 80's into early 90's. There were tons of PC makers out there, IBM became dominant but they haven't been in that business for a long time now. Microsoft was the enormous winner coming out of that.

When grandma starts talking about GPU's and NVIDIA stock, this AI bubble might be at its peak. I think if history rhymes with past lyrics (that whole history doesn't repeat but rhymes thing), its more likely that a software company (OpenAI, Google, Microsoft) is going to be the dominant player emerging from a potential bubble burst.

A potential pin-prick to the bubble (looking into my crystal ball): pricing for AI - once companies realize their runways are running out and they want to turn a profit of some kind to stakeholders they will need to raise prices to offset the energy and hardware build outs. Customers may realize they aren't getting the productivity to offset those increased costs and hold back budgets on AI spend. Or chip tariffs depending on TACO volatility.

And note: it took Cisco 25-years to recoup its market cap bubble from its high of 2000.

https://finance.yahoo.com/quote/CSCO/

Isn't that true for other companies that survived the times? That the bubble created extreme highs before it burst and now the companies have grown more naturally?

For example, Intel: https://finance.yahoo.com/quote/INTC

Honest question. Not my area of focus.

The difference is that Cisco gear uses industry standard open protocols that anyone can duplicate. Nvidia’s current moat is CUDA.

But Amazon, Google, Tesla and to a lesser extent Microsoft are all creating their own chips to lessen their dependence on Nvidia

This.

I am still thinking and trying to figure out what the new concepts i.e social media, streaming, etc, of the AI era will be. There are of course the extrapolations of existing concepts, e.g shopping just like it adapted to the internet world will adapt to the AI world and you'll have a few companies there -- equivalents of uber, amazon, zomato, etc (or maybe those companies themselves).

Products with no margin such as translation will of course get a million times better, but I mean in terms of economic value.

5% of my social media feed is AI generated. I can easily imagine a completely AI generated feed that is custom tailored to my interests. The future is already here. It is just unevenly distributed.

> is there going to be an impending crash leaving only the ‘best and brightest’ on the other side?

The way it usually goes is that the prize goes to best real life application. It's not entirely clear what that is just yet. I was involved in web commerce in the 90s and there was a good bit of casting around looking for applications. And not really any mobile use, though I did a bit of that too. It's hard to visualise the future.

There's a persistent pessimism - almost to the point of paranoia - in tech from the trauma of the dot come bust. People fear waves of mass layoffs and 99% reductions in the value of their companies.

In reality most things go in an S curve and the excitement fades in a rather boring way. People were worried about bubbles in 2016, but they never really happened.

There was never a mobile bust, or a social bust, or a big data bust. These things just got boring after a while and the valuations stabilised. Sometimes companies went bust and sometimes they kept growing, but the super exponential growth in validation stopped. And we moved on to talking about the next thing.

The way people tend to invest is compare it to a different number. I remember one point where FB was valued around $100 per user or so. The IPO flopped when people thought it was too high, but it turned out far too low. Netflix, Amazon, etc, had a similar situation too. Logically it was overinvested, but they broke the records.

People were cautious after dot com. Then they got greedy with FAANG. Even greedier with crypto. Crypto proved that markets didn't really matter.

AI actually has some basis, but some of it is being measured to crypto standards. Not to mention it was already overinvested 10 years ago when everyone was calling it "machine learning", and yet the numbers kept going up.

Sometimes prices are just high though. Bubbles happen when someone is trying to sell it to another sucker, who's trying to sell it to another sucker. We don't see a lot of "sucker" language going around with AI yet. When it happens, people will advertise it as a great investment, like they did with NFT and real estate near the bubbles.

I was too young at the time, but people who were there at internet companies in the 2000s what was it like? Did people think the same way as today where there is too money going in and not enough value coming out? Was it obvious to insiders only or to the general public as well?

These days it seems even a significant porting of the general public is aware of the overhype. But back then internet wasn't much of a thing so information didn't spread as fast so I imagine a lot of people didn't even know there was a huge hype around the internet.

It didn’t last past 2001 (lasted from 1996ish to 2001) and much of the money was stock and not real money so the paper millionaires never had the money since it never vested by the time they were hired until the time the dot com company went bust. VA Linux was a prime example of a company that didn’t have a profitable exit and evaporated with the market correction that I remember started in late 2000. The 9/11 attacks in NYC erased any chance of a dot com recovery.

Everyone knew there was this thing called the internet because cable broadband was being deployed everywhere and being sold for less than cost. That company whose name I couldn’t remember went bust I think around 2003(?) and I think AT&T took them over and doubled the price.

I remember someone showing me that you could do to domino’s pizza website and get a free pizza (unlimited orders) just for going to their website. I think that was 2002-3(?). There were lost of giveaways during that era. It all came to a stop as the internet grew and costs pilled up more so than the dot com bust.

No, it is not the same at all.

I worked at a small investment firm in college during the time and I remember the portfolio manager doing the portfolio allocation for a retirement plan. Randomly picking Janus mutual funds in the hall right before he went to lunch.

I remember the secretary complaining in December 1999 how she knew a secretary at another firm that got rich from an IPO.

There has just never been a level a speculation like in late 1999 that I have ever experienced. This was over geocities level webpages on dial up modems with almost no business plans to even make money.

I do miss it somewhat. At the time, a selected number of us on my product team were given retention bonuses because the company didn't want us being poached for internet jobs. I'm not complaining, because it was free money but anyone wanting to poach us away would probably have just offered more.

OTOH, I remember being in SF in spring 2001 for Embedded Systems Conference and overhearing a couple of Web developers on the same bus talking about how everyone they knew was losing their jobs. So I guess by that time, the end was in sight! Our bonuses continued for at least a year after that.

This is the first time I ever hear of temporary "retention bonuses" in big Cos to avoid poaching.

So I guess the hype was worse back them because I don't think anyone is getting them now haha.

Yeah, once they were sure that the whole .com thing imploded, the bonuses went away. Then we used the lack of bonuses as leverage for raises :-)

FuckedCompany.com was the source for tips about this. We'd check it every morning, and it predicted our company's layoffs like clockwork.

I was working in Atlanta GA for boring old profitable enterprise companies and the environment didn’t change at all.

I started working in 1996. By 2001, I was making $70K. The company I worked for did bill printing for utility companies. To put that in perspective, I had my first house built in the suburbs for $175k. This was well within the rule of thumb of not spending more than 3.5x income for a house.

On the other hand, I didn’t have that much invested in the stock market. The entire dot com bust was a shrug and curiousity to me while it was happening.

On the other hand, the real estate crash in 2008…

While all of my latest projects (cloud consulting) have in some form another involved “AI”, they are really just AWS API calls that are much easier using Gen AI LLMs. But could have been done with much more development and training of ML models before.

When the AI bust happens, it will have the same affect on me - none. Until businesses stop needing people who know how to translate the latest tech into business value, my career is safe.

It was largely the same. A ton of Internet companies struggled to find a way to make money after the initial land grab, and most failed. The hype was breathless and over the top.

A lot of late 90s to early 2000s hype has come true. Today we have substantial companies with no physical office and that have never used physical mail, delivery for everything, and ubiquitous connectivity for devices. It just took about 15-20 more years to get there.

I suspect we will have AI replacing programmers on non-trivial and non-slop projects in 15-20 years.

The social optimism of the late 90s and 2000s about the Internet was mostly wrong though. We didn’t really anticipate either surveillance capitalism or mass disinformation breaking whole segments of the population off from reality. Like most well intentioned people we had trouble even imagining the uses that “dark triad” personality types would have for the technology.

The pessimism around AI is, I think, an overcorrection from the excessive optimism around the net and the web.

It felt like we all had an equal chance at success back then. Now it's limited to those with enough money or very rare luck. The riches of today were sown almost in the first decade of the 2000s, it seems. Who today will get rich from AI starting from nothing? (not that wealth is the end goal, but in the context of this thread it is)

I find LLMs incredibly useful for some very specific cases: summarizing text for example. Even dialogues to learn something.

Now there are n use cases (10 <= n <= 50?). There is no real moat.

What we don't know is, if there will be significant advances, and how useful they will be. But we also don't know that for any other area.

What we do know however is that breakthroughs are rare.

Everybody is now in the hype train. AI here, AI there. I find most of it just annoying.

My guess is this that a reckoning is more probable than the new super feature.

Most AI companies are losing money like the dot com era. They spent a lot on buying chipsets or training models, or paying an unsustainable rate for API access to LLM companies. https://ethanding.substack.com/p/windsurf-gets-margin-called If AI tools are charging the world the cost it should be, then homo sapiens will be way cheaper in most of the cases.

Personally, I don't believe they are the same. I view AI as the next natural update to worker productivity like PC and windows. It's like the old days where people put MS Office on their resume as a skill, prompt engineering or whatever AI is gonna be exactly like that.

This feels like a very realistic take and is probably already occurring given what you can see on LinkedIn and the like: people are using “prompt engineer” in their job titles ancillary to their primary role.

Like every industry, there are thousands of competitors sharing the same niches in its nascent years, and will gradually round out to a small number of incumbents.

The biggest risk for any startup is they depend on the core model developers, who are wisening up and designing their own tools to undercut those who just call their API's.

You can game this mathematically. Compare historical rates of investment in AI against AI revenue and return on investment. Then once those lines are graphed do the same for the dot com bust and compare the two resulting graphs.

If the two graphs so far look the same you have your answer. Ultimately this will not be a finance problem but a market and behavior problem. People are predictable with confidence.

A bubble sort of implies this ephemeral quality that is just super wrong.

I think its better to describe AI as a big shake up. Like a Cambrian explosion of sorts.

We will see tons of bankruptcies, in both AI and not AI companies. A ton of low profile ones and likely many high profile ones. It's just the nature of it.

Most new businesses fail, and any large shake up in an industry will create new winner and losers. I don’t see AI as being any different. If it will be a crash or more controlled will likely be a question of scale and other factors.

While I think its similar, I don't think it will burst quite the same. I think we will see companies rolling back, but not all companies, and not all at once.

Yes.

Maybe someone should make AI-infused monitor stands.

/s maybe

I bet you could make deck slides of this idea as a joke and still find someone to write you a check.

I don't think it will burst but more likely deflate.

Year on year people will be disappointed that AI isn't consistently improving and that it will eventually plateau, and other things will be hyped up excessively.

I can imagine quantum being the next big craze but at the same time, I doubt that can even get to consumer considering it needs near 0 kelvin temperatures.

But who knows

We won't see anything on the scale of the dot com bust again, but certainly the market will correct. All markets on a long enough time line are going to weed out the failures and leave you with the 'best and brightest' - that's a feature, not a bug, of capitalism.

With AI, the question isn’t if there’s going to be a bubble… it’s which startups will still be alive when it bursts. Good luck with that.

Yes I think so. Most big tech companies are American and very over-valued compared to what the profits are. The American stock exchange is worth about double that of the European. I think the American stocks and companies will see a downturn in valuation and European and Asian companies will see an increase.

Other sectors that people today find is boring will have an upswing.

Not remotely. It’s not even logical.

From an employees standpoint, most of the employers weren’t profitable and when they went out of business, employees weren’t having a hard time finding a job.

But honestly, it was only people working for “tech” companies. I was working in Atlanta as a Windows developer with four years of experience working for a boring old bill printing company. Other profitable enterprise companies were still hiring like banks, insurance companies, airlines, etc.

Today, all of the major tech employer’s are trillion dollar profitable companies with real businesses that are not dependent on Gen AI. They have been using machine learning forever and still will.

Every company doesn’t need to be a software development shop. But every company doesn’t need to be into data analytics.

On the other hand, from the investor’s standpoint, the only publicly traded company that will live or die by AI is Nvidia - and Tesla which has always been a meme stock. If investor money is lost in the private market, who cares?

Yes I know that there is a push to allow private equity investing in 401K plans and pensions. I really only have sympathy for pension fund holders who don’t have a choice.

AI is going to create endless jobs by allowing everyone to make a huge mess.

We are currently in the early phase where people are creating code messes with little oversight and future people will be expected to build on them, which will either take expanded effort or complete re-writes from scratch.

The quality of all products will suffer for the foreseeable future.

It's now a several times a day occurrence for me of trying to help someone figure out a problem, and when I ask them why they did that, the answer is always that the AI told them to.

If the people that are using these tools don't know right from wrong, it's a recipe for disaster.

This problem is amplified in non-technical fields where people are creating tools for businesses without even the knowledge of what a code review is.

AI will be around forever, and the problems it will create are immeasurable, but unfortunately, it will likely always be seen as the solution to the problem as well.