When Tailwinds Vanish | Hacker News Reader

When Tailwinds Vanish

luttig.substack.com

104 pointssethbannon35 comments

Interesting read. Not sure I agree, though not my area expertise.

The idea that we’re hitting full saturation, or that startups will have to push into market vs pulling sales from market gaps assumes that we’re reaching full saturation across the board. That the reach of everyone being online also means they’re all online in the same markets / products / verticals.

I’ve worked most my career consulting as a software engineer, primarily split in 4-5 business verticals. The giants in those spaces (hiring, healthcare, cosmetics, etc) have 8-15% market share. And there’s constantly turn over where people for whatever reason exit and return to the customer pool.

So while from an economic and academic standpoint this seems pretty solid, I have a hard time accepting it practically.

Again, anything I know about this kind of thing is peripheral knowledge, so very likely wrong. would love some insight :)

Interesting article, but I disagree with the author on a lot of points.

it largely revolves around: "But people can’t spend more than 100% of their time or money on the Internet."

first off, although the internet is baked into everything, its not just about the internet. self driving cars, 3d printing, fake meats, and lots of other things I can't imagine.

2ndly, while you can't always take more time/money from a person, you can always add more people. there are something like 2 billion people living on 2$ a day. Its not inconceivable that the market/economy today is .1% of what it will be in 30 years.

even before covid19, businesses that were not taking advantage of the internet and tech was dying such as Macys... Businesses that dont invest like crazy into tech and R&D will go extinct.

I dont think the internet is maturing, its just getting started.

also: "The Silicon Valley of tomorrow will not look like that of today"

The SV of 1999 looked not much like 2004, which looked not much like 2009, which looked not much like 2014, which looks very different from 2020. Its not stating much, everyone knows that. It also furthers the point, there is a lot of growth in the future.

Was about to say the same, in particular this:

> also: "The Silicon Valley of tomorrow will not look like that of today"

I came to live in Silicon Valley in the 80's. At the time, the technical press was bemoaning that the glory days were gone, making chips had become a race to the bottom, engineers were being laid off, and there wouldn't be enough jobs for the people who were already in the valley, much less the people in school who would graduate in 1 to 5 years with engineering degrees.

It bummed me out because I was at Intel, which seemed like a good company when I joined them out of school, but they were laying off people and it was "clear" that the semiconductor houses were going to consolidate down to nothing.

At the time I took a "big risk" and left an established company and joined a startup that seemed to be doing okay. At least they weren't losing money. It was called Sun Microsystems, they had just gone public (the Friday before I joined), and were less than 2,000 people. They were building these things called "Workstations" which everyone figured would fail because as soon as the "Big" companies (HP, IBM, and DEC) decided to get serious about this space it would squash them like a bug. They were doing some crazy things, like telling you all of their APIs and letting anyone copy them. They were using computer chips they bought from Motorola and running a University's version of the UNIX operating system on them. Their marketing message was simple, "We're open, we may not be here next month but don't worry because everyone has the information they need to make compatible stuff for our systems so even if we're dead, you can still use our gear."

It was a very different message than what the "big boys" were saying.

10 years later they were the "gorilla" and DEC, HP, and IBM were having a hard time keeping up. Then the Internet caught on an suddenly they were in a maelstrom of innovation and change and couldn't change with it. Ex-sun folks like to joke they became the dot "." in Oracle.com.

When the dot com crash hit I was older and had gone through the semi-conductor crash so I wasn't sure that this was "the end" as everyone was predicting. There would never be another VC who would be willing to risk giving you money. There would never be another retail investor that would buy and IPO for a company that wasn't making money. There was no "moat" or anyway to protect your market share so the only companies that would survive would be the distilled and conglomerated husks of the old dot com dinosaurs.

Then Google and Facebook exploded.

So what does it all mean?

It means that if this is the first time you've seen the phoenix burst into flames you will just know that its dead and there isn't any left to do but sweep up the ashes. But if this is your second, or third, or fourth, trip around the block you know that the last trip found a match between value and technologies ability to deliver it with enough margin to support building a company. And you know that people are going to spend what they earn and there will be new things to do and new ways to spend. If you can see the things that made gave the market wings over the last 10 years, often there are hints about the next thing that will generate the excitement.

And if you can't, that is okay because the person that can? They are going to need talented engineers to help them bring it to market and make it work. There is something magical about taking $0.02 cents of electricity and turning it into something someone will buy for $1. People that can internalize the value of information will understand how to capture value while building new products. And that in turn will fund an entirely new ecosystem of businesses, large and small.

There is nothing more creative, and innovative, than a group of engineering friends with too much time on their hands.

This is a great take, and really puts things into perspective. Thanks for taking the time to write it.

I’m with you. Every generation seems to have a time when the “it’s all been invented” meme takes off. There’s plenty of innovation in our future. It just takes imagination to envision it. When I hear people saying otherwise, I see it as an indictment on their powers of imagination.

Thank you for sharing your experience, that was insightful.

The long view of history, rising and falling waves of innovation, funding, market demands, new ideas and shifts in technology.. After each winter, a new cycle begins with a search for:

> a match between value and [technology's] ability to deliver it with enough margin to support building a company

Which sounds like "product-market fit".

That makes me think, what may seem like a downturn in the industry or saturation in a market segment, could be a sign that fresh pastures are needed, with new opportunities for those who can see where things will go, the next big ideas (which often start small and look like a toy to the established/entrenched players).

> If you can see the things that gave the market wings over the last 10 years, often there are hints about the next thing that will generate the excitement.

This is what I love about the world of computers and technology. The people who inhabit it, who push it forward, are driven by imagination and a kind of undying hope, to see the blue skies beyond the weather of today.

I remember in the early 70's being told that with the end of the space race, there was not going to be jobs for electrical/electronic engineers!

> Its not inconceivable that the market/economy today is .1% of what it will be in 30 years.

I kind of followed the argument until this part. You seem to be saying that a 1000x increase in market/economy is possible in 30 years, mostly as a result of the number of people being around and productive. This doesn't hold up, IMHO, unless some other (massively more dominant) multiplier is factored in.

If we can make 2 billion people 10x richer than they are today, market/economic growth is not necessarily capped at 10x. Systems don't have to be linear.

As a contrived example, I would imagine that folks living on $2/day probably buy on average effectively 0 iPhones/year. If somehow they had 10x the wealth they do now, they wouldn't buy 10x the number of iPhones they could previously buy... they would buy almost-infinity-x the number of iPhones they previously could buy, thanks to division by almost-zero.

BTW, my point is not that this is some kind of arithmetic trick, but that 1000x growth just depends on how low the starting baseline is (which is very low, by almost all measures of baseline economic activity/wealth of the $2/day demographic).

I agree that systems don't have to be linear, but for it to be non-linear, it'd have to feed back somehow. Simply buying X times as many iPhones doesn't do it, unless e.g. those iPhones allow them to be Y times more productive, etc.

My point of view goes something like this: some parts of the world _are_ developed, and these already have the various additional productivity multipliers we theorised about above. These can be the baseline productivity level for our calculation. So if the rest of the world were to be brought to this level, the end result is indeed linear, and more or less depends on where we drew the line for "developed". However, it is not 1/1000th of the world population. We could be talking about a 10, or maybe 20x multiplier if we're very generous.

I also don't think historical precedent favours such optimistic numbers, though we could be living in interesting times. However OP's point was, again, a simple introduction / upliftment of more people, not technological multipliers.

Also economic growth is still tied to finite resource consumption, so 1000x the economy would devastate the earth. Sure, humanity's economy may become decoupled from finite resource consumption, but that would be a profound change in life on earth.

There was not that much tailwinds in the rest of anyway, SV will probably become more like London or Berlin.

What is more interesting to me is that the Internet is no longer magical. It’s not more fun that the electricity and everything revolves around few apps and services that produce the same content over and over again. Tiktok is the only new big thing in years and it’s not from SV.

The interesting things happen in the finance and news media. Right now these suck and innovations are happening there. There’s many things that suck too but nothing interesting is happening on most of these.

In other words, the internet as we know it is about finding and connecting what’s out there and that’s almost complete. The internet of tomorrow is about making sense of what is out there and restructuring it.

I think there’s a lot of green fields for tech just around the corner. But unlike the previous boom that one will be heavily connected with politics. It’s probably going to be just as destructive but less constructive and less fun.

One underlying assumption behind the article: information technology stands still.

The big boys and behemoths of current and past eras emerged when there was a major change in technology or adoption. The two are also interdependent. New technologies drive new use cases and adoption in new segments, which drive further technological inventions.

Ubiquitous wireless devices, new AI capabilities, and wearable sensors are some areas that provide opportunities for startups today. Many more are still to come in the next few years and decades.

Some technologies I keep an eye on: Adaptive Robotics for everyday tasks, Autonomous Vehicles, Brain-Computer Interface (BCI), Quantum Computing

They are still far from practical, but hold great long-term possibilities.

Feel free to add some that interest you.

Distributed on-demand manufacturing is something I am following. With different types of what you might consider "No Tooling Change" manufacturing methods getting better and better, we might see the economic viability of manufacturing on demand in local manufacturing centers. These depots could make items asked of them on the spot with no upfront re-tooling costs or bulk orders needed. This could enable small businesses to have novel ideas that can easily scale across these near instantly available manufacturing nodes, avoiding hefty tooling and bulk buy costs, and also avoiding having to ship things across the globe.

Imagine you order SomeDeviceX from coolthings.com, they send the order sheet to the town's local fulfillment center in a standardized format they accept. The fulfillment center sends the files out to it's manufacturing devices, think CNC, 3D printers, laser cutters and PCB laser etchers. Some time later the device is assembled, ready to ship out to the neighbourhood.

What I like about the idea is that some items are valuable to only some people, so could never achieve economies of scale that make them viable to a business, and never be manufactured cheap enough locally to meet the market at the right price point.

The only kind of manufacturing being done in many places is large scale plants for things outside of the community, and everyone relies on products shipped in from around the globe. There is a gap missing, between expensive boutique items and mass produced cheap items, where affordable locally produced items could exist if manufacturing could be cheaper.

As a heavy R&D type, I enjoy these trends, though suspect it's predicated on a big and historically wrong IF: no new compute platform.

mainframe -> pc -> laptop -> phone -> cloud/gpu/ai -> blockchain/vr/... -> quantum/self-driving-car/... .

Each of those bring massive high-margin software tides to ride.

In practice, life's still pretty sweet for cloud/gpu/ai startups, so we still seem to be roughly at that point. (We do a lot of gpu cloud accelerated security/fraud visual graph analytics: most enterprises still do it on-prem and basically just hope splunk can run regex without dying.) In GPU land, it's pretty clear they're being way underutilized, esp. due to slow SW rollout by various incumbents like cloud vendors.

We're still early on the transition to the next phase. The blockchain/vr/quantum startups raised a ton of funding, but have largely fallen flat. Some of the fundamentals in them seem sound, so they seem more of a matter of being early, not necessarily wrong. Current teams just couldn't get the tech together, or a true killer app: those are a question of time. Likewise, we know quantum etc. are coming, just need ~5 more years for what most general sw people start to get into.

IMO bigger risk is monopolies. Maybe the market reasoning is more from VC perspective, where everything feels 'expensive', and SaaS metrics era made investors forget how to take tech risks while the coming gen of platforms emerge.

EDIT: Some other potential near-term up-and-comers - synthetic bio (crispr,..), NN moving into program synthesis (so past perception and into automation), serverless. I don't follow agtech, but drones/robots/etc. are increasingly here.

Except that VR and "blockchain" were duds.

Self-driving cars ought to be big, though. Sooner or later, someone will get it right.

The Next Big Thing is probably wind and solar + batteries. Bye bye, gasoline industry. In twenty years, you may need to plan out a route in advance to be sure of finding an open gasoline station. Heating oil will last longer.

> Except that VR and "blockchain" were duds.

Based on limited experience, albeit going back about 30 years, VR won't become popular before visual and proprioceptive inputs can be synched. And I doubt that'll be practical before direct brain input works well.

About blockchain, it'll succeed once users don't know that it's there. I mean, how many understand how HTTPS certificates work?

Funny you mention HTTPS certificates because there's a blockchain-based protocol that's trying to create a decentralized certificate authority. It basically aims to shift the root of trust from CAs to the blockchain https://handshake.org

"direct brain input" Please no ..

I mean, I think the key difference is that HTTPS certs are actually useful.

Orchid creates multi-hop VPN connections using servers from multiple providers. It uses a private blockchain cryptocurrency to ~anonymously distribute value from users to VPN providers. That's useful because users don't need to trust any single VPN service.

Loki is an onion network. Currently the Session messaging app is the only service implemented, but I gather that there will be others. Loki uses a private blockchain cryptocurrency to hinder Sibyl attacks. That's useful because users will less likely get pwned by malicious nodes.

If those work out, there's no reason why most users would need to even know that blockchain cryptocurrencies were involved.

VR has gone from 0.3% to 1.5% of all users on Steam in one year, and still rising. I'm old enough to remember people saying that both mice and gfx cards weren't worth the money and never would be. Dismissing VR as a "dud" is... well, foolish :-)

VR isn't "3D TV". It's only going to become more commonplace as the tech and UX improves.

Self-driving cars have been five to ten years away since ~2005. I've pretty much lost hope that it will ever go much beyond "nicer cruise control".

> In twenty years, you may need to plan out a route in advance to be sure of finding an open gasoline station. Heating oil will last longer.

Huh, why? In many countries you're not allowed build new homes with oil heating systems, and there are plenty of government scheme to get rid of existing examples. And of course in highly urbanized countries heating oil barely exists anyway (it hasn't made sense for a long time if mains gas is available).

Found some of the points in this article to be contradictory. Author argues that carefully measured growth will win in the years to come, but also suggests that companies need to go all in on SG&A since having more staff means more growth.

Personally, I agree with the carefully measured growth part, but I also think small teams still have a lot of power. Investing in SG&A still isn't going to be as fruitful as R&D. The Internet hasn't halted yet – there's still a lot more to uncover. The physical world gives us a perception that everything is finite – that everyone's fighting over the same resources. A zero sum mindset (which would certainly make the case for more capital spend on SG&A). But in the virtual world of the Internet, the universe is ever-expanding and new ways of doing things are being discovered at every corner.

This is an incredibly blinkered view. Even if you think the US is saturated, which I doubt, there's a literal entire world out of there, many parts of which are still taking their first baby steps onto the Internet. Places like Indonesia and Vietnam are growing their "internet economy" (money spent online) at 50+% CAGR, and COVID is accelerating many (not all) companies in spaces like e-commerce or online collaboration.

https://www.temasek.com.sg/en/news-and-views/stories/future/...

The question then becomes is SV the best place for technology to be built for India/Malaysia/Vietnam or others. My guess is that the answer is no, and after the success of China we'll see protectionist policies that ensure that the answer is no. A lot of the power of US companies comes from US branding, which I think still holds at the individual consumer level. But at the country policy level, is the US still the gold standard that developing economies should aspire too? It was pretty clear to people living in the west that cracks have been forming for a long time, but the COVID situation has enough propaganda potential paint various government structures in a different light.

Unfortunately the parts of the world that are now coming online are mostly poor countries that won't increase revenue by much, at least for a few decades.

Aside from the other great points that have been made already in other top comments, I have a little nitpick (or feedback for the author, if they're a reader): charts should have axis labels, units, and ticks. Otherwise you might as well not have them.

Take the second chart on the page for example, as soon as I saw it a little warning light turned on in my mind. That chart says incremental spend (y axis) vs time (x axis), but no values or ticks on either axis. A chart like this could literally mean _anything_ because you have no idea of the rate at which those lines are getting smaller. You have no idea how much more revenue to startups there is compared to incumbents, the difference that _looks_ big in the chart could even be 0.001%, just "zoomed in".

That article seems to have some /very/ bad advice shifting away from R&D. Did they not remember what happened to everybody else who did that? Eventually overtaken when they got comlacent. They all thought themselves the controller of matured markets or the emerging ones as a completely separate domain. They thought wrong.

There is no guarantee that R&D will get thr same raw returns or even stay on top of the market but neglecting it from trying to not cannibalize their own markershare has proven a fatal mistake to companies.

Even being an "also ran" is better than getting your lunch eaten when the next new thing comes along.

Looking at this from a FOSS perspective, some of the "negatives" this article discusses strike me as significant reasons for optimism.

Collaborative, community-driven, open source projects sometimes struggle to be as innovative as startup executing a singular vision – but FOSS projects excel at providing compelling alternatives to existing, entrenched platforms. If the pace of new "breakthrough innovations" slows a bit, it could provide FOSS exactly the opportunity it needs to catch up on features before users move on to a new paradigm.

I think there's room for growth.

There's still a big part of the growing world population who doesn't have internet access.

Internet connects not only people but also devices.

However, the growth of online businesses will mean shrinking elsewhere which I am not sure it is good. Amazon and the like will replace brick and mortar stores, Netflix will replace TV, food ordering will replace restaurants and so on.

There will be less time spent outside, less social interaction and the jobs created in IT industry might be less than the jobs lost elsewhere.

Am I the only one confused that this might be about the CSS framework?

I've never heard of it. The meaning of the title was clear to me. The title wouldn't even make sense without modification if it was about a CSS framework.

Reply on news.ycombinator.com