Comment by happy-go-lucky | Hacker News Reader

Comment by happy-go-lucky | Hacker News Reader

The essay itself is general and surprising, and the combination never loses its effectiveness.

Ideas keep evolving. There was nothing new about the idea of a search engine when Google came into existence in 1998.

In 1945, Dr. Vannevar Bush wrote an article titled As We May Think:

https://www.theatlantic.com/magazine/archive/1945/07/as-we-m...

> he urges that men of science should then turn to the massive task of making more accessible our bewildering store of knowledge.

In the 1960’s, Gerard Salton created and developed Salton’s Magic Automatic Retriever of Text (SMART). He also authored a book called A Theory of Indexing detailing his initial tests that search is largely based off of relevancy algorithms. Here’s a very interesting blog post titled Search Down Memory Lane by Tom Evslin who worked with Salton during this project:

http://blog.tomevslin.com/2006/01/search_down_mem.html

Fast forward to 1990, a man named Alan Emtage created the first search engine known as Archie which retrieved a database files by matching user queries using regexes. Following the growth of Archie’s popularity, two similar search engines Veronica and Jughead were created and they started indexing plain text files.

In 1993, the first bot called World Wide Web Wanderer was created and then it was upgraded to capture active URLs and store them in a database. Then came ALIWEB (Archie-Like Indexing of the Web), which crawled meta information of pages.

There’re others including WebCrawler (1994), Lycos(1994), AltaVista(1995), Excite(1995), Yahoo(1995), Dogpile(1996), Ask Jeeves(1996).

This is just my superficial understanding of how the idea of search engines was born and has evolved over time. They were very different from Google, but they’ve similarities to how data is processed and analyzed today.

- Facebook wasn't the first social network. People were connected, shared pictures, and had friends before.

- Amazon wasn't the first online store.

- AirBnB wasn't the first short term rental website

- Netflix wasn't the first online movie portal.

- Apple didn't make the first cell phone or even the first smart phone, or the first music player or the first home computer.

- Tesla didn't make the first electric car.

- etc.

Still, the one criticism that startup hear again and again is that "You are late - it's already there".

There's value in being first to market but what those example really show is that first isn't all it's cracked up to be, and marketing and sales are crucial for success.

I’d say all of those examples show that you need to be first to the actual market, not just first or first to _some_ market

First to market is a tactic and potentially an advantage, but it isn't the end all be all of whether or not you'll be successful.

Just to play devil's advocate, what was the first online store? I think Amazon was pretty damn close, having started in 1994 as a store that simply stocked a large selection of rare books. The shipping took weeks as I recall.

It was not an obvious idea at the time, simply because there weren't that many people on the web. It was around the time that SSL was being developed [1] -- so if you were much earlier you probably couldn't take payments securely.

There were probably mail order catalogs that had websites, but I think that is a different thing than a "web store".

Also, Netflix wasn't the first online movie portal, but I'm pretty sure they were the first people to make a business out of mailing DVDs! It's crazy that they started like that.

[1] https://en.wikipedia.org/wiki/Transport_Layer_Security#SSL_1...

Vaguely pertinent, but ViaWeb was founded in 1995 by pg (and others). Amazon was apparently founded in 1994, but wasn't called Amazon until 1995 -- or, at least based on my cursory skim of Wikipedia.

So, I don't know that it was the first or not, but Bezos was certainly not the only one thinking about it at the time.

> I think Amazon was pretty damn close, having started in 1994 as a store that simply stocked a large selection of rare books. The shipping took weeks as I recall.

Afaik, they didn't really carry any stock when they first started. They were re-selling mail order books.

A while back there was a documentary about Amazon's early years. The part that everyone always quotes is the thing about using doors as desks. (Which struck me as silly, doors are expensive compared to many large, flat surfaces.) Anyway, the part that really stuck with me was that they had one favorite mail-order place, except that shop had a ten book minimum order. So Amazon would order the one book someone wanted and then nine copies of some obscure book about snails that was never in stock.

Yup I think I saw that same video on YouTube. They didn't have stock, so they were sort of like a meta-retailer. They presented an illusion of large selection.

I think the value was the interface and ease of searching, and not inventory and retail operations.

CD Now beat Amazon by a few months: https://en.wikipedia.org/wiki/CDNow

But you're right, Amazon was basically there at the beginning of online retail. A better example is YouTube. Many online video sharing sites had started and died years before YouTube was a gleam in Janet Jackson's pasty http://usatoday30.usatoday.com/tech/news/2006-10-11-youtube-...

I think YouTube hit the right combination of UI and network speed improvements. Before that, either the UI was glitchy or the average person's network speed was too slow to stream video usefully.

Also, I believe the only way to deliver that content to a wide variety of browsers at the time was Flash. In fact Adobe/Macromedia may have only added the video functionality in the prior 5 years or so, so before that I think you couldn't even do it without making people install a plugin.

Remember this was well before Chrome came out, and Firefox was still nascent. So a lot of people were using IE.

I think Flash was already starting to fall out of fashion by then, but they realized there was a valuable technology deployed in it (video codecs).

Both true, Amazon was one of the first web stores, Netflix was one of the first movie portals. And similarly, the ipod and iphones weren't the first music players or smart phones but the first ones that worked really well. Remember how you could chose between 20 different mp3 players, all quite cheap but usually in crappy quality and without a music store? Remember how you could buy several smart phones like an Ericsson with a stylus or a BlackBerry with a keyboard?

Doing it right matters more than doing it first.

Is there a distinction between "its already here" and "its already being made popular by someone else"?

That distinction could be demonstrated by market maturity, which could be further defined by how large the incumbents are and how saturated the market has become.

Google is a giant now, and has made the search engine market both mature and saturated (along with its competitors). Before Google came along, there were several other search engines, but the market wasn't at all saturated. It was much easier to enter that market with a well placed innovation in searching while the market was primed to grow even larger among a set of competitors who had not dominated it yet. The market was large enough that an innovative company would capture first time users and spread itself organically among them, eventually stealing them from other engines in the process.

Conversely, the search engine market has become saturated. To start a new search engine, you'll need to steal users from an existing search engine, because there is virtually no one who doesn't already use one. As a basic concept, it's extremely difficult to differentiate on the core idea of a search engine. You will not beat Google just by offering a new technical innovation, you'd likely have to beat Google by adding search innovations to a novel platform that is emerging to challenge Google before they capture that platform as well. That sort of combination would offer a new fertile market that hasn't been saturated. As an example, a search engine for applications would have been a legitimate threat, had Google not already developed it for Android and Chrome with Apple competing.

Right, but I was only disputing OPs point about "its already here" being an illegitimite concern about a startup company.

The future is a spiral of iterative improvement with occasional leaps up a few rungs on the spiral.

Google didn't make the first search engine, either.

What general ideas did we sow in the past 20-30 years that, through minor iterations, are coming to (surprising) fruition today?

Machine learning

Yup

You can trace the origins of the ANN to the 40s and 50s. Since then, statistical ML has come into vogue a few times, but not like today. Phones and other data sources (plus the algorithms to properly deal with all the data) have ushered in our current renaissance.

Huh. Why do you single out phones as a data source? That seems like a weird statement to make. Neural nets are in vogue because everyone has a phone?

Machine learning and neural networks are in vogue because it's easier than ever before to collect data for them to be trained upon and process. Phones are a part of that for sure.

Everybody has a camera in their pocket that takes pictures for free. That's increased the number of images that you can train a neural network on (and the number of images that you might want to classify with it) dramatically.

I remember when I was a kid, you had huge binders of photo albums, and the total number of pictures that a family might own numbered in the hundreds. Now I take that many photos on a week-long vacation, or sometimes just in ordinary life, and my cloud backup has maybe tens of thousands. Multiply that by a billion cell phones and that's a lot of images out there.

Data was never the issue for neural nets. It was the compute power required to run them. The most popular image recognition dataset today was mostly taken pre-smartphones.

When I was in school, studying ML, we were always able to beat NN performance with SVMs on the datasets available. It's only through larger training sets that NNs start to shine.

NNs will still be beat by svms on simple datasets, even big ones. Where NNs shine is when the data has an underlying structure that can be exploited. Like how convolutions take advantage of the 2 dimensional nature of images, or RNNs take advantage of sequential data. Algorithmic improvements like dropout and proper weight initialization have also made them bunch better than they used to be.

If you know what the structure is, then you could write a proper kernel and then SVMs are better again. NNs win when you don't know what the structure is and it's too complex to approximate with a standard kernel.

I'm not aware of anything like a convolutional SVM, but I'm not terribly familiar with them. As I understand it SVMs are fundamentally shallow learners, and if you try to hack them to make them deep and recurrent or convolutional, you just get weird NNs.

Shallow in the sense that SVMs are not layered, but the layering of a NN is only to enable the humble perceptron (logistic or sigmoid, usually) to model more complex functions. In contrast, the SVM doesn't need to layer, because the kernel can be arbitrarily complex.

Meh Duda and Hart came out in 1973. I could just be somewhat old and bitter.

Free unix.

Cryptocurrencies.

Military robotics.

Personal drones.

Augmented reality.

Freedom of information.

Pervasive media piracy.

Darknets.

State level hacking.

Technofascist dystopia in general.

Increasingly cheap wind, solar and battery hardware.

Smartphones (also enables vastly cheaper batteries and drone components).

The idea of search engine wasn't new but even today millions of linear algebra students are amazed when they find out how google/page rank works

How does it work?

https://en.wikipedia.org/wiki/PageRank

Really neat, simple, iterative algorithm.

Reply on news.ycombinator.com