Frighteningly Ambitious Startup Ideas (2012)
paulgraham.com
paulgraham.com
Over time, hardware gets cheaper, and the cost of building a search engine declines. This has not been matched with a reduction in ad density. The cost improvement is going into Google's profit margin. That indicates a vulnerability.
Google has some striking weaknesses. The two I focus on are provenance and business background. I've done work on business background (see "sitetruth.com"). That tries to find the real-world business behind a web site that's selling something, and then uses the data available about real-world businesses to check it out. That can then be fed into search result ordering. SiteTruth is a demo; it's running off free data sources. Paying for higher quality data from Dun and Bradstreet and other non-cheap sources would make the business background check much better.
Google's other big weakness is provenance. Search engines should find the original source of information. Much of the Internet is sites scraping other sites, linking to other sites, and commenting on other sites. We see this on HN all the time, where someone links to an article, but the actual source is two blogs deep. The original source should be the primary search result, perhaps annotated with notes about the more heavily promoted sites mentioning it. This means more attention to when something appeared and better matching of content. Google already does this for news. For ranking purposes, attention metrics for scraper sites need to be credited to the original source, not the scraper site.
One possible customer for such a system is Bloomberg. The people who get those expensive terminals could use both of those features. Then offer it to the better universities, so that students grow up using something less consumer-oriented with Google. That strategy worked for Facebook.
I rather believe, smaller and more niche search engines like zomato(for restaurants in india) etc would get better of google very soon. Like, imagine a search engine for online learning resources with filters like beginners/intermediate/advanced, books, courses, etc; that would be so much better than what google has to offer right now given it indexes a lot of quality resources.
This is all the more true on mobile, where the search is meant to initiate an action: buy something, make a reservation. Being able to complete a circuit in one place gives competitors in some domains a powerful advantage over Google.
Something like Alexa, or other voice assistants, are probably closer to what that Google killer would end up looking like. Good thing that Google got in there early.
Killing another search engine is exactly how Google got big.
There isn't anymore. It needs to be fundamentally different.
I rarely visit sites out of a band of 15 or so, and I only use Google to find pages within those sites (which itself follows a normal distribution). I.e. I know I want a page from stackoverflow, or Wikipedia, or arxiv, or Reddit, or Yelp but rather than navigating to those sites I just cmd+T in Chrome, type keywords, and scroll to the site's link on the result page which is almost always at the top anyways.
I'm sure others use Google like this too (varying on your culture's go-to sites), so does it even need to exist as it does right now? I could imagine a new app that indexes just within this band of sites. And Google or some other app is relegated to "the rest" for rare occasions.
Add interesting sites and resources in your profile and get your search system. Now it's public beta, you can add unlimited resources, but now 1 link - 1 indexed page. And we working on deep indexing system. We have crawler, but we need improve our platform, architecture and other.
On our platform now >150.000 resources, that handpicked and added by users.
I think when you are designer and add 1000 best resources for design - it can be better search system for this topic.
And you can provide your opinion for ANY topic. It can be more transparent, flexible, democratic search system, where people select best resources for search. And where every opinion is important and can be different.
Personally, I find duckduckgo to be an excellent option; it does almost all of my search for me, with none of the privacy issues.
if you make it your default search, you can revert to google by using the banged g operator (g!), which will anonymize and then submit your search string to google. very seamless; very convenient; more privacy.
disclaimer: no personal affiliation with ddg
Therefore a competing product doesn't need to be as good as current Google, only as good as it used to be. Still hard, but far from impossible for someone with money.
What may be harder is Cuil's second problem as mentioned: Finding a revenue model. You can't compromise your data by manipulating searches (lessens quality) and you can't compromise your integrity by prioritizing ads (lessens trust). I'm sure a lot of people would pay big money for the answer, which no one seems to have (as demonstrated by Twitter's descent into ad trickery such as camouflaging the ad disclaimers).
Perhaps the answer is to do what Craigslist did. In a few categories, ones you have to ask for such as "Shopping" or "Jobs", there are ads, or product ordering boxes, clearly identified as such. Non-commercial information is not polluted with ads.
Try and search anything even remotely technical or specific. Google brings back a few thousand irrelevant results, often with stuff like 'not found: word you were actually after' underneath in grey. Sometimes even when you do add quotes or speech marks around the term.
A better Google would be one that doesn't assume people want 'any old possible answer' and tries to answer specific queries as easily as general ones.
While there seems no barrier to entry, just consumer habit, google is faster than it used to be, despite a larger web and more users.
Speed of search results has been shown to be extremely important to users. Google's many large server farms, close to users, is a huge capital investment. Google has doubled down on this advantage, with google suggest, which requires even more power. For example, google has much lower latency than DDG. Replicating those server farms would be expensive; I'm not sure than even all of AWS could do it.
When I first used Google in 2000, I immediately knew they would come to dominate search, because their product was so much better. It's not enough to be as good as Google, you have to wipe the floor with them.
As others have pointed out, the Amazon Echo is a viable approach to doing this. Shame it sucks for anything but the most trivial queries right now.
Actually, I appreciate that feature very much. I like that when I search for Python, Google knows as a programmer I mean the language not the snake. Doubtless there are herpetologists who appreciate that Google knows they mean the snake.
Page and Brin were right on the problem and the solution for the time.
While I want my startup to be still more successful, I'll thrilled if it is as successful as what Page and Brin did.
However, my understanding is that Google CSE would basically replace site search and can be configured to operate identically, and isn't slated to disappear.
- environmental degradation is based on two exponential growth trends: human population and per-capita consumption (driven by economic growth). Solving either could be, literally, a save the world scenario. The first is easiest technically, (healthcare, non-coercive birth control, education, economic opportunity for women) but a political minefield. (yes, population is still growing exponentially)
- dealing with climate change. At a minimum, this means figuring out how to handle a significant rise in sea level in the next 100 years. Walls? Relocation? We need to think about response now, more so than avoidance.
- actually doing geo-engineering to prevent/reverse the greenhouse effect
- preventing a nuclear war
- preventing a large asteroid collision
- dealing with large-scale disruption of employment (for example retail continues to be hammered by online stores, self-driving trucks, etc.) Build a better Mechanical Turk might be a useful strategy.
Obviously some of these are low-probability short-term events, but they have very high costs. In a sense, you would be taking advantage of people's tendency to systematically under-invest in problems of that kind.
Arguing about how to express that is sort of pointless.
It is in fact a logistic sigmoid function. It is amazing, in that it is not what you'd expect : clearly the vast majority of species have the ability to correctly estimate the carrying capacity of a habitat and only overshoot by a little bit.
So the Malthusian argument is incorrect, probably, if humans are like other animals (and since population growth curves in many localities do indeed look like S-curves I would say that this ought to be the default claim). Human population size will by itself stop just shy of the carrying capacity of the planet. Which is exactly what you'd want to happen.
In all likelihood, this will happen without active measures from governments (which won't work anyway due to the global nature of the required policies), with or without popularizing contraceptives, with or without female labor participation, ... (not saying those aren't desirable, just that they aren't necessary)
The alternative requires an assumption that humans are hugely different from the species that we evolved from, and since this certainly isn't true in the biological sense, why would our population growth be so very different ? Humans are not the first large animal to "conquer the planet", and reach population numbers in the billions, with some fish well on the way to trillions even today, and we aren't the biggest species in terms of total biomass by a very long way (amongst non-plants that honour likely goes to ants, but even pathetic species of bacteria dwarf their numbers). Will we be the last one ? Perhaps, but it seems unlikely.
https://en.wikipedia.org/wiki/Logistic_function#In_ecology:_...
It's very likely that humans exhibit the same behavior, but on a micro-scale, this isn't very comforting. If a politician says "Well, the earth will be fine, but 10% of you are gonna die and another 30% won't be able to have children that survive to adulthood", there would be rioting on the streets ... which has a non-negligible chance of bringing about those numbers.
Nature will be just fine, but remember that the mechanisms by which nature self-adapts are the four horsemen of the apocalypse: pestilence, war, famine, and death.
> The alternative requires an assumption that humans are hugely different from the species that we evolved from
Well we are, for proof look at the fact that we're even discussing it here when all the other species never did. Look, that all sounds very peachy if it weren't for the fact that our species alone has grown the brains and the self-justifications for consciously attempting to continously outwit "nature" (and be able to replicate way beyond what our ancestral habitats could carry) at every turn with agriculture and all the sciences =) granted, there's no point in or way of "going back" for all manner of reasons, but we needn't lose sight of certain basic facts here..
https://ourworldindata.org/world-population-growth/
Growth is clearly exponential. Your theories don't fit the data.
Can you explain what I'm missing ?
Look at the linear function. It experiences "decrease in the rate of growth" along the line.
If the exponent is not constant, we do not have "exponential growth" anymore — by definition.
In this case G is the difference between the birth rate and the death rate for time period t. As long as a G is positive, you get exponential growth. Every time. This is how compound interest works. The fact that your balance next year is based on your starting balance + the balance multiplied by a positive growth rate ALWAYS leads to exponential growth in the balance.
Here's a video that explains it. https://www.youtube.com/watch?v=NB4KVSjgi24
OK then. Let the birth rate be 1 person per second and death rate 0 (nobody dies). Do you get exponential growth? No. This is still linear growth.
Your another mistake is that ax^b is exponential growth, and ax^c is also exponential growth (if b and c are constants), but these are different processes. If you get a single process with changing exponent, this is no longer exponential growth, it can be arbitrarily defined.
One person per second is linear.
.00001 person per year * current population is exponential, assuming only that the (fractional) person is added to the population factor so that the base is bigger next year.
If you don't understand that, you shouldn't be commenting on growth rates.
If you assume the growth rate is exponential, then you will see that the first year, the growth rate is 100%, the second year the growth rate is 50%, the third year the growth rate is 33% and so on. At any point, you can fit an exponential to the graph tangential to the current point. It won't fit much, but you can still do it.
A decreasing exponential may be exponential, but it also very well may not be. But we'll need to look a few years into the future to be sure. That said, expert opinion seems to be that growth rate is at this point linear (or iow an exponential that decreases each year so as to appear linear) [0]
[0]: https://ourworldindata.org/wp-content/uploads/2013/05/update...
No, a "decreasing exponential" is not exponential. It might return to an exponential after the decreasing period, but during the period of decline, the growth is not exponential.
Look at [0]. 2016 had a higher population than 2015, yet the growth was smaller. Not just a smaller percentage, but actually smaller by about half a million people. If growth was exponential, then the higher population would have a larger increase, but it doesn't, and so the growth isn't exponential.
[0]: http://www.worldometers.info/world-population/world-populati...
You are describing the shift from a steep exponential growth curve to a lower exponential growth curve that I mentioned earlier.
If you get 10% compounded interest on your money it grows exponentially at 10%.
If you get 9% compounded interest it grows exponentially at 9%.
And if you _were_ getting 10% and the bank changes the rate to 9% your money is now growing exponentially at 9% interest, assuming only that the interest is reinvested.
Then going from $100 to $110 that would be 10%.
Going from $110 to $120 would be ~9%.
Going from $120 to $130 would be ~8%.
In any year you could conclude my savings grow exponentially. 10%! 9%! whoohoo! But if you look over a longer time frame, it clearly isn't.
And as nevdka mentioned, the population growth is less than linear. If my bank gave me $10 of interest last year and $9 this year, and $8 the next, there is definitely no exponential growth.
A growth curve that is, in fact, logistic, would look like that.
It's been continuously declining for half a century after peaking, the assumption that it's suddenly reached a bottom at the most recent measurement and returned to an exponential steady state is unwarranted.
So how do you factored in all the possible coming or not coming atomic/biological/chemical wars + new breakthroughs in medicine/agriculture?
And since right now "More than 100 million people worldwide are facing acute malnutrition and risk starving to death" [0] I would call it a bit cynic to say "Problem solved".
[0] http://www.independent.co.uk/news/world/politics/number-of-p...
Additionally, more humans are expected to die from obesity than from hunger, as of now.
I did not say that. I just said that the problem is clearly not solved, when so many people are starving.
One might expect that access to food and medicine leads to population growth, but in the long term, countries with the highest levels of access have the lowest population growth. It's quite a strange phenomenon, and it is basically this trend which leads statisticians to estimate 2070 as having a peak population.
Yes, but it might be a missleading one. Because I think our society causes many people to consider children as expensive and time-consuming and therefore not wanting to have children.
But that might change again. There were times when childrens were be seen as a luxery or as a necessity, a insurance for old age.
I mean, I am not saying it will change to that or that direction, but I think you can't really predict the future on how the world is acting today as human society changed quite drastic in the last hundreds years. I would rather predict that human society will continue to change and therefore I am very cautious with any fixed predictions.
E.g. I can certainly envision post-work economies, where the "penalty" of having multiple children is much lower, as they are essentially guaranteed a good life. I wonder which direction society's value system would evolve in that scenario.
Because if you're talking about the whole world, I'm pretty sure you're incorrect.
People need to recognize that this is, currently, just a trend, and not yet a fact and certainly not a physical law. For example, large families are correlated with religious fundamentalism, which is rising. Could it rise to the point where the trend changes? It is possible.
It's not a physical law, but still pretty well understood phenomenon: more likely your offsprings are going to survive, less likely you are going breed more of them.
Rosling also claimed that population growth projections have been very accurate for decades.
For example, what is erdogan thinking when he says Turks should have five or more children? I'd love to be able to read what's going on in people's heads. Personally, I couldn't give two (insert expletive) if my "culture and heritage" were to disappear or dissolve I to something new in fifty or a hundred years.
You left out one detail: environmental degration per consumption. This is a really important factor, and is probably easist of th tree to minimize, and is being minimized every year. (Assuming that you didn't equate consumption and environmental degration).
I'm cautious with this: If you know what problems people/society has, you know what products to solve/mitigate these problems might be bought. This perfectly sound like startup opportunities to me.
In 2004, the United Nations' "High-level Panel on Threats, Challenges and Change" analysed threats and challenges to international peace and security. Their top ten list in priority order: Poverty, Infectious disease, Environmental degradation, Inter-state war, Civil war, Genocide, Other atrocities (e.g., trade in women and children for sexual slavery, or kidnapping for body parts), Weapons of mass destruction (nuclear proliferation, chemical weapon proliferation, biological weapon proliferation), Terrorism, and Transnational organized crime. [1]
Jean-francois Rischard's "High Noon: 20 Global Problems, 20 Years To Solve Them" also identifies 20 issues broken down into 3 groups [2]:
Group one (sharing our planet) includes global warming; biodiversity and ecosystem losses, fisheries depletion, deforestation, water deficits, and maritime safety and pollution.
Group two (sharing our humanity) includes massive step-up in the fight against poverty, peacekeeping-conflict prevention-combatting terrorism, education for all, global infectuous diseases, digital divide, and natural disaster prevention and mitigation.
Group three (sharing our rule book) includes reinventing taxation for the 21st century, biotechnology rules, global financial architecture, illegal drugs, trade-investment-competition rules, intellectual property rights, e-commerce rules, and international labor and migration rules.
edit: formatting.
===
[1] https://en.wikipedia.org/wiki/High-level_Panel_on_Threats,_C...
[2] https://www.amazon.com/gp/customer-reviews/R28KHPSYSQ7O6T/re...
GOOG: we've got a search engine that we did as an academic project. Lots of people are using it. Let's turn it into a business.
FACE: we've got a social network that lots of college kids are using. Let's turn it into a business.
AMAZ: let's sell books over the internet.
AAPL: Hmmm, my crazy genius friend Woz built an awesome "personal computer" in his spare time. Let's scale up production and sell it.
MSFT: Let's scale up our software development business and go after much bigger targets.
Also worth looking up: wishful thinking.
Larry Page was incredibly ambitious even while he was a PhD student. He naively intended to work on 3 projects. Luckily he listened to the good advice of his advisor, Terry Winograd, who asked him to narrow his focus down to the web.
The quote below is excerpted from his 2009 commencement address [1].
"I think it is often easier to make progress on mega-ambitious dreams. I know that sounds completely nuts. But, since no one else is crazy enough to do it, you have little competition. There are so few people this crazy that I feel like I know them all by first name."
[1] http://googlepress.blogspot.com.ng/2009/05/larry-pages-unive...
EDIT: Statements like "survival bias" are used by people that often subscribe to Narrative Fallacies.
Companies like GOOG, FACE, APPL, MSFT are not successful because they won some big lottery.
They are successful because the identified "objectives" and made the right calculations to those objectives.
And if they stop making the right calculations to their ojectives they WILL end up like Kodak, Nokia & HP
What matters is whether you are ready or not. To make use of your "Luck"
Bill Gates, for example, initially declined IBM's approaches to build DOS for them. He declined, and referred them to Gary Kildall. IBM only came back to Gates because the deal with Kildall fell apart. There was no vision, skill, etc, in that sequence of events. Just luck. I imagine other highly successful people could point out similar stories in their past.
Gates likely would have been successful either way, but perhaps not to the same degree.
They were more swift in executions, standing up to "conventional" wisdom, and persevering when others were running away.
This is pretty close to tautological. Your version of "right" can only be evaluated decades after the fact. Who succeeded? The ones who were right. Who's right? The ones who succeeded.
I also think it highly overvalues making the right calculations. All the calculation in the world will not avail you if you are not in the right position to make it happen. In human history, there have been about 100 billion people. How many of them, even given the right ideas, were placed to build a company like Apple or Google?
Even if we focus on ideas alone, making all the right calculations isn't sufficient. Somebody else could do that, too, and then it does come down to circumstance and luck.
It's certainly uncomfortable that life is hugely random. But I don't think the solution is to try to sweep that under the rug. If nothing else, that sort of aversion to risk keeps us from making good calculations about business.
I think this list is a good argument against survival bias.
Google had a fantastic product, which really worked. They tested on a small number of users and found it to be fantastic. They scaled that. There is a clear logical explanation for why they succeeded.
You could say "well, they may have failed at scaling". How hard is it to scale from where they already were, with such a good product? Not very hard, I'd say.
Let's assume your argument holds, and that many other groups had the same idea and didn't succeed. So they must have gotten at least as far as Google in getting their product done and tested by a decent sized group of users. So maybe 100s or 1000s of people were using their product on a daily basis. In which case, why don't we hear about those products?
I think you can treat survival bias seriously when there is a large element of luck involved. In the Google story, I don't see much luck. I see a really good idea that worked and was then brought to market.
Lots of people think they have fantastic products.
Lots of products take off without being fantastic.
Lots of products can be really fantastic but never manage to take off.
Google didn't think they had a fantastic product. They _had_ one, it was pretty self-evident that the search engine was a great deal better than any competitor, and they had empirical evidence that many other people agreed (as they were using the engine).
> Lots of products take off without being fantastic.
I'm not sure that's true. I can't think of any examples of terrible products that have been hugely successful. I can think of examples of products that _became_ terrible after success or monopoly had been achieved.
> Lots of products can be really fantastic but never manage to take off.
It took me less than a minute using Google to see that it would conquer the world. Same for my colleagues. I think really you mean "good". _Fantastic_ products are by definition super-useful, popular, have instant appeal.
Most succinct and accurate apple history I've ever read.
We might well have seen the last web/consumer tech behemoth that was 'started in a basement' too. The market is simply far too concentrated these days. The ability to quickly disrupt a behemoth certainly isn't a permanent state of affairs for our industry.
If you look at it that way, there are always emerging areas that have little competition because they are so difficult. So if you had the ability to and were smart enough to create an innovative product in your basement, I think you absolutely could grow it into a behemoth.
Starting a computer company today would be very difficult and costly. Starting a computer company at a time very few people understood them or knew how to build them would be a lot more feasible if you were an expert (Woz).
Companies get disrupted by companies that are almost by definition starting in a seemingly completely different area.
OILCO: Hmmm, we've got this strange sticky black flammable liquid bubbling out of the ground. Let's sell it as a fuel.
According to Mrs. Packard:
"Well, it was very scary to be leaving General Electric, and to realize that we were on our own. Actually, he [David] came back to get his engineer's degree, so really we were thinking of ourselves as students for the first year."
"During that year, he and Bill started Hewlett-Packard. At the end of his year's leave of absence was when he had to write a letter actually quitting General Electric. That was the most traumatic part of it all."
GOOG: Hey lets save the entire internet on our servers and be better than anyone else by magnitudes
FACE: I want to have everyone in the world to have a profile on facebook
AMAZ: I want every book in the world to be sold on my website, and pretty much everything else.
AAPL: I want to be better at product design than and become better than Microsoft
MSFT: I tricked IBM into doing distribution for us, but I want a PC in everyone's home within 10 years.
I think those come closer to the founders visions than what you wrote. Some of my versions actually include mindsets that their founders at the very beginning.
That's why words can influence people so heavily, even though the facts are the same.
For example Paul Graham states you may be looking to replace email with a todo list like application. But all you really have to say is you're building a todo list application. He says this prevents investors from saying are we there yet and likely thinking you are a bit too ambitious.
Not that there's anything wrong with that, and PG is right to point out that the rewards for doing so are huge (this reads a lot like Zero to One).
But in terms of social impact I look at these ideas, with the possible exception of Moore's Law (albeit they are from five years ago) and think, meh?
How much does an incrementally better Google, a slightly faster GMail (exists, if you're willing to give up spam filtering), a replacement to Oxford/Harvard, a slightly better Netflix really get us? Certainly it doesn't seem to solve any of the major challenges that we face today.
- Google: Actually there's a lot to improve on Google, and a ton of value there. Just think that instead of what Google gives on medical info, you'll be refereed to an AI that will ask you questions, and give you top quality information like a medical expert would(or better - more time, better quality content, etc...). Or the million ways better search could help engineers/scientists/lawyers/etc...
Actually i would like to see something like a subscription marketplace for quality search engines(where your query is refered to the most suitable search engine and profits are shared fairly). Would be interesting.
- Replace universities: Just solve the affordability problem. That would be huge enough.
- Ongoing diagnosis - If it works well, and affordable enough, could create plenty of value. But that's a big if.
But the rest, sure they seem mostly about money.
I'm also fairly convinced big tech companies will eat up any emerging market that's within it's competency unless there's some core violation of principles. Take Duck Duck Go, for example. They've definitely managed to steal market share away from Google because they clash with Google's lack of concern for pure privacy. Snapchat managed to do the same to Facebook because of their disappearing images, but now that Facebook has incorporated those features into Instagram, that point of difference is dead.
For as long as big tech maintains the startup mindset, challengers will invevitably be swallowed either through competition or acquisition. In my opinion, the most ambitious startups are those pursuing underserved markets encumbered by bureaucratic or corporate incompetency.
"Ambition" is the wrong word here.
I think a better search engine at least does the following:
1. Focuses on access to technical information. It doesn't need to have up-to-the-minute indexes of reddit, or hollywood whatever magazine. As an example, lets suppose I want to know how to design an HVAC system for a house. If I google "home HVAC design" I get a page full of HGTV installation videos and HVAC company brochures. One result had a single math equation, and a different result had a data table. If I "Better Search" for home HVAC design, I should get almost exclusively technical information, such as equipment lists, equations, and data tables.
2. Allow personalized scraping. If I want to extract math equations and data tables from all websites related to "home HVAC design" this should only take a few clicks. Give the user tools to OCR images from results, extract data from graphs, and import code snippets into their IDE.
3. Call out StackOverflow. Lots of times the answers to technical questions use information that is out of date, or cite solutions/workarounds from webpages that cease to exist. A better search engine would identify situations (not just on StackOverflow) where this happens and put warnings next to any results that make heavy use of dead websites or old information. A better search engine might automatically suggest cached versions of the now-broken links, or suggest that you resubmit the question to whatever community and ask for more up-to-date information.
What does all this do? Hopefully it dramatically changes the productivity of technically-minded people by making research easier.
Edit: really cool that And it looks like it these are being attempted! There are companies doing the diagnosis in the early stage, there is slack, there is duck duck go, etc. Dont see the second Jobsian hardware anywhere though, and don't expect to.
There are some decent counter-arguments:
The draw of universities, certainly top ones, is that you physically bring together hundreds, thousands, of people who are eager to learn. This is something you will never get with a MOOC in the same way. You might have a community ten times larger doing 6.00x online, but you don't have that mingling of smart people in different fields. Good ideas (rarely) come out of lectures, they come out of meeting friends for beer or coffee.
That's a very rose tinted view of reality, but some universities like Oxbridge, MIT, Stanford, etc really excel at putting great people in the same room as each other.
Aside from education, universities are by far (in my experience) the best places to meet like-minded people who are a similar age to you. We had societies for everything under the sun. Many people go to university/college solely for the social aspect.
Many have tried to fight Google based on the (incorrect) idea that having "a better search engine" will win. I guess this is why people still keep building these things.
You can't win Google by "building a better search engine". Many smart people before you have tried and didn't succeed, and it wasn't because they were not smart enough, or because the product was inferior. It's because search engine has become a stable category, just like you eat the same cereal every morning.
People won't suddenly ditch all cereals from general meals just because some newcomer cereal company makes a "better cereal". You have to build something completely different, or at least position it that way so people perceive it as a different category.
While all these people were "trying to beat google", it was actually Facebook that came close to "beating" google at its own game. Don't just build a better something. Build something different.
Edit: There is an exception to your understanding of perception. Extreme convenience can overcome market perceptions...If someone found a way to make a search engine or solve discovery in a way that is extremely more convenient than using Google, they can unseat the brand.
Can you say where the scariest and most ambitious convincing pitch was on the following scale? 1) We're going to build the next Facebook! 2) We're going to found the next Apple! 3) Our product will create sweeping political change! This will produce a major economic revolution in at least one country! (Seasteading would be change on this level if it worked; creating a new country successfully is around the same level of change as this.) 4) Our product is the next nuclear weapon. You wouldn't want that in the wrong hands, would you? 5) This is going to be the equivalent of the invention of electricity if it works out. 6) We're going to make an IQ-enhancing drug and produce basic change in the human condition. 7) We're going to build serious Drexler-class molecular nanotechnology. 8) We're going to upload a human brain into a computer. 9) We're going to build a recursively self-improving Artificial Intelligence. 10) We think we've figured out how to hack into the computer our universe is running on.
Turns out it's not about new search engines but rather how to capture search intent. Amazon did that to beat Google.
2. Replace Email
This will be done by notifications and direct messaging. Kids these days don't bother with email.
3. Replace Universities
Turns out this is just upscaling. Most of the job training stuff is dropping out of the lower orbit of a university and being swallowed up by the new in faux DeVry tech boot camps. But want to be an AI expert? Want to be a doctor? You still need to put in time in academia.
4. Internet drama
They all mostly caught up. Still can't beat Netflix though.
5. The Next Steve Jobs
Conan O'Brien, a fellow Harvard alum, said it much better. In the pursuit of being someone else you end up being yourself. Just be yourself.
6. Bring Back Moore's Law
This is happening in GPUs.
7. Ongoing Diagnosis
This is happening.
3-5 years is about a cycle that turns high tech to low tech. Most of the stuff talked about in 2012 are now low tech. Things like mobile app development or tablets or 4k TVs. Twas a good feeling back then, now it's all AI and MR and rapid fabrication.
This seems to be a life phase, not a long term trend.
Everywhere else you have a company email account and it's your principle way to have a paper trail.
"Not using email" is definitely a kids phase. Then you go work for a non-startup company and email is everything.
An ambitious startup idea would be oust Microsoft Outlook from this role - find a way to globally distribute the meeting room asset tracking stuff securely and seamlessly.
That's one reason for every lawyer's advice on corporate emails: anything older than at most 2 months must get automatically deleted, even when there is no interaction from you, preferably without exceptions. Things that require longer term storage should simply be printed out and stored offsite. Anything you find offensive, anything about your employment conditions, ... should likewise be printed out and go to offsite storage.
This is not destruction of evidence because you're doing this as a standard policy "to save space and to declutter" (hah !), not as a reaction to being sued. As to why that argument convinces anyone when large companies can reasonably be expected to get sued very regularly ... But it is pretty well established in legal precedent.
You can download the emails of course, but then the accusation that, firstly, this is against your contract (it probably is forbidden in your contract), and secondly the emails could be faked will definitely be there. And of course, policies are likely in place to prevent you from doing this, even if they're very unlikely to be effective.
I felt like that little detail was very much missing from the conversation here and yet I feel it is very important.
What does my paper trail get me? Evidence to be spun into a scenario that justifies firing me if the wind ever blows the wrong way.
One of the companies I am talking about has revenue in the billions and operations across the globe and is publicly traded. The employees, young and old, want to communicate in real time, to get things done quickly. Many employees even have chat access to the highest levels of the company, including the CEO.
Wow, all moving to proprietary, incompatible solutions. What an improvement kids are these days! It sounds like we are back in the 80s of computing. Thanks, I'll keep the email I can use virtually everywhere.
This is very much doable nowdays for cheap. Hmm, I should stop slacking on my hobbies and roll one specialized search engine for something that I think will gain traction.
Google's results are not trustworthy at all. Will there be a buy option at all behind a Google search result ? Good guess. Shipping available/reasonable ? Odds are not good. Not a scam ? ... not sure how to judge that one at all ... and of course there's no shortage of scams on the internet, some of which rank pretty well on Google.
Amazon is not optimal though. They're not reliable enough for my parents for instance. I guess the issue is that compared to a local store they still suck.
Still compared to Google and webshops, they're definitely the superior product.
Making predictions on what kids do is really, really bad. "hey my kid only cares about drinking and going out with friends", does not mean that he'll become a drunkard when they become adult.
> This will be done by notifications and direct messaging. Kids these days don't bother with email.
Pg is not talking about personal emails for messaging. Personal messaging is mostly "read, reply, and forget", and you are right that this is now better served by instant messaging protocols (modulo the fact that there is no universal IM protocol).
He is describing work emails, as a "todo list" as he puts it. You get such emails from your colleagues or business partners. They generate work for you and you cannot just reply and forget about them. You need to work a long time (either doing research or the actual work), and get to that email thread later. It is a "shared" todo list, and you send messages to synchronize the state.
Google's Inbox is built exactly for this purpose, although I believe pg wants something more revolutionary.
Have you recently bought a car through text messaging? Lots of people do that now. They text four or five sales reps at different dealerships, they all run off to do work, then they come back with the best prices and you negotiate entirely through that channel.
Have you recently bought hair extensions through text messaging? Lots of people do that, and entire field sales force are managed and communicated to with text messaging.
Do you work with 25-35yo somethings? Guess how you're getting most of your work instructions these days. It's text messaging.
Have you seen the additional UI/UX changes? Do you know about how a big group of ex-Apple employees rolled off to spin up their own text messaging company that extends the purpose of SMS with additional functionalities that's now more and more common place in-app before we see it roll out to the first party channels.
Email is fax. Tech grows up. Keep up when you can.
Your examples are bizarre. Buying hair extensions is a tiny niche market to begin with. The percentage of people buying cars from dealers via text message is a rounding error, and I highly doubt that they would have been using email to do it a decade ago, so it's completely irrelevant.
Text messages and similar are fundamentally no different from email except that they have a few limitations that limit their use cases (length, subject lines, attachments, etc). If those limitations are removed...you basically have email again.
— A platform for CRISPR-assisted human genome editing. I, for one, welcome improving our inefficient, disease-ridden, short-lived bodies.
— A morality API. Various human societies have different moral concerns and taboos. Can we expose some tweaks and tools to navigate this as continuous spectra?
— Post-government transitional structures. Hierarchical power arrangements and traditional governments are doomed anyway; let's use network science to simulate and explore how things might work in the future.
And I still want my thermonuclear power generation. If electricity is to finally totally replace combustion, we need it cheaper. MUCH cheaper.
The "big problems" being solved in tech right now are mostly just people fiddling around with tooling and business models. I wish more of us would dig into these kinds of opportunities... things with a profound possibility to change civilization for the better.
I think there's a lot of skepticism about that kind of endeavor, because there are so many 20 year olds who think they are going to "change the world" without either understanding the world, or really having the deep seated motivation required to see such a project through hell.
But the fact that the area is blanketed in charlatans doesn't mean the underlying concept—looking deep into our future and try to accelerate big difficult changes—is flawed.
It's just the people best suited to do it—with the skepticism, experience, and skin in the game—tend to write it off as a silly idea. Maybe they need to do more drugs.
I'm actively working on #2 and #3, which I think are probably the same problem in different forms, but I think my efforts in the area would be well characterized as "hampered by mental illness". Not to say I won't get there, but maybe you have to be a little crazy to try.
I asked HN how to deal with this but my question was too difficult to read since I've tried to make thought experiment of imagining you are back in 90s.
Here is my post: https://news.ycombinator.com/item?id=14336513
May be somebody can give thoughtful answer to this deep question.
You can think about your place in abstract historical eras on your next mushroom trip in Mexico.
In retrospect, this seems to be a wrong conclusion. I know Bing is far far away from Google; But I think that Bing investment was well worth it for Microsoft. If you have followed Build 2017 conference, it is obvious how Microsoft is leveraging AI/ML in several of their products; and of course, they have a good offering of cognitive service cloud APIs. Without their heavy investment in Bing, Microsoft perhaps would have found it hard to make such quick jump onto the AI/ML bandwagon.
As a side note, is Microsoft still considered a company on its way to oblivion? I personally think they reinvented themselves in the new era. Am I mistaken in my view?
You are not mistaken but back in 2012 when the article was written it was pretty common to bash Microsoft.
In fact, PG had written an older essay with a click-bait title of "Microsoft is Dead" [1] in 2007. Naturally, many people didn't realize he meant that metaphorically so he wrote a rejoinder titled "Microsoft is Dead: The Cliffs Notes" [2].
>The most ambitious is to try to do it automatically: to write a compiler that will parallelize our code for us.
Does anyone know if this has been accomplished in any form or fashion by anyone/startup?
or is there anyone doing the below part atleast?
>An optimization marketplace would be a way to generate the sufficiently smart compiler piecemeal
I am very interested in finding a solution to these problems.
> Does anyone know if this has been accomplished in any form or fashion by anyone/startup?
There are compilers that are able to auto-parallelize specific kinds of software fragments, but most of them are not able to optimize general-purpose programs in a reasonable way.
There were and still are research projects in many computer science departments that try to find ways to auto-parallelize C or Fortran code, but many believe that these low-level languages provide a too unconstrained memory and type system, which makes auto-parallelization efforts really hard, because a "sufficiently smart compiler" has to prove certain properties about the code to guarantee correctness of the optimization (parallelization).
There are also people who think there's not much room for parallelization in general-purpose software development (business applications) that mostly wrap specific libraries. The libraries itself may exploit parallelization, but this is usually hand-crafted by their developers and not done using tools.
I am not saying, it's not possible, but it has been tried many times, and building a new language (or improving one) may be less effort, than trying to auto-parallelize classic, old code (C, Fortran, etc.).
But I think it is an event protocol of sorts, with TODO being a subset. Subject, msg body, etc. are universal to a lot of things. For example, the difference between say email and SMS is more UI than the actual structure of the messages. The text transcript of a phone call could fit into the same bucket.
To point is, the "email" UI / UX is awful. It hasn't evolved. It might as well be a fax machine. (Goggle) Wave might have been a step in the right direction. The idea was right and solid, but Google (being Google) fucked up the launch / rollout. And now we're still stuck with the same shite email experience. And thus Graham's article is still relevant.
Note: We're all but stuck with the email protocol. The world isn't going to change en masse. That is my client experience can be improved without having to abandon the protocol. Come on. That's not risky, nor is it rocket science.
Email doesn't tread well. That is, it's only good for focused single topic conversations.
Email keeps passing a copy of previous msgs back n forth, instead of centralizing a single copy.
Wave was a step in the right direction, and Google lost its nerve, and vision. Foolish.
Yes, it's being extended. But a screwdriver as hammer isn't typically a best practice.
Personally, Im like to have something that integrates email, SMS, social messages, etc into a single pro-sumrer friendly experience. A given conversation can, and often does, span the various mediums, but it's still a single conversation. When someone sends me a message with a link I shouldn't have to look in multiple places. It's 2017. Is that too much to ask?
My father is a big guy, so I think about what I eat and I walk to work.
For most medical stuff I agree that an ounce of prevention is worth more than a pound of cure. But we don't need to create more tests and we certainly should think through creating businesses that want to create revenue from it. There is an insurance industry we need to contain.
You have the inputs, you have the outputs, now you can perform analysis to improve lifestyle health.
People are not calorie counters, they're intrinsically lazy, they'll mostly accept the defaults their fridge and cupboard tell them to do.
I think you could save millions of lives this way.
It's one of the elements in 1999 AD, the Ford corporation video made in the 60s, but it's never been implemented.
Surprising that there aren't more, given that it's literally the #1 cause of death
Is the hypothesis that hackers are more likely to spread the search engine to other users, or is it something else?
A Mac Nano for $99 to compete with the Raspberry Pi.
A Mac Mini for $199 to spread the love of OSX.
An iPhone mini for $199 to spread the love of iOS.
A 60" Apple TV (yes, with screen) with an iPod like remote.
Siri.com as a text search engine.
Apple Pay as money transfer between individuals.
If Apple wants to keep their high quality mark untouched, they should spin a second company codenamed "Orange" and deliver lower quality devices in order to spread their platforms.
It is time for Apple to leave the constrained niche they've cornered themselves into. 10% of the top market is a good position to hold forever, just like Rolex, but in the era of social virality market penetration is a must for survival, you must be everywhere and in every device.
Things an end user might constitute a critical modification for some market, like a bottle opener super glued to the back of a phone, is a blemish in the eye of Jony Ive. Apple employees are grossed out by hacks, so they can't operate in the long tail. They can only go after the big glorious spike toward the perfect unified One True Way and therefore are destined to hold only 20% market share.
They serve that 20% damn well though. People who want designers to think things out for them.
There's another kind of designer, who designs for chaos, for modification, for play. But those people don't work at Apple.
For payment networks total penetration is crucial, for open markets, social platforms, productive software, for search engines. Again, a Rolex can survive as a timeless jewel in your wrist since it does not need interaction with the rest of the world, but good luck competing with Whatsapp, Facebook, Amazon, Google, Paypal, even Uber or MS Office if only 10% of your userbase can afford your exclusive products.
That's why I propose they create a sister business in charge of expanding their market penetration with cheaper products or face extinction by isolation. Apple for the elites, Orange for the plebes, both using their platform which in turn will allow them to capture a greater market share for their future technologies.
I can't do Facetime with my mother or brother because they don't have Apple products. We could if they had an Orange product for $99 running OSX or iOS.
I'm sure even in 2012 some of his "$50 ideas" (or no, he said he'd spend $1000 potentially!) could be solved with the existing solutions and a little tweaking. I know for a fact gmail has a reduced loading mode that is snappier and lighter weight.
> Is there no configuration of the bits in memory of a present day computer that is this compiler? If you really think so, you should try to prove it, because that would be an interesting result.
I'd again invite the author to "try" and recreate a human mind in computer memory - surely "some combination of bits exist" for it, and attempting to do so "would be an interesting result!" etc etc
What this basically reads as is a bunch of "here are some simple things that can be reinvented by oneself, but I'm lazy so someone make new services here so I can come in and comment about how I was just thinking I needed one and that I'll 'check it out' and..." For the time that's passed between this article's writing and now surely he could have just hired a hacker with that "$/month" he'd otherwise spend to set up a configuration that better suits his whim (likely from existing tools, userscript tweaks, settings/labs/etc) and pocket the difference.
It's extremely difficult to get into an existing product category such as operating systems, search engines, email, etc.
I firmly believe the next wave of technology will hit our TVs. The final nut has been cracked and that is the remote. Typing on a TV is the worse and voice won't help, so a dual touch/keyboard piece of glass is all needed to make the TV the next entertaining device after the mobile phone. But they need to converge all devices needed right now into one. I don't need a pc, a mobile working as a remote, a tv device and a tv screen in order to enjoy the future. I need ONE TV with a glass remote. That's it. Google and Apple are missing the elephant in the room.
What would Steve Jobs do? A 60" Apple TV with an iPod like remote for $3995 in dark aluminum and a 40" version for $1995 and still sell the attachable box for $150 for those who already have a shitty TV. Allow Safari and all apps without restrictions, even Popcorn Time. Can't fight what the consumers want.
i'd like to be able to search google for "all the pages that contain a particular javascript snippet", for example. obviously, google has that data. obviously, they dont want to make it available.
there are services that offer such things but they suck because their datasets are way too small.
on google specifically: the results pages dont matter anymore. nobody clicks anything but the "#1 results" if they click even that. for questions, google very often responds with something like a quora question or some widget of its own, or something else that is really useful - and if its not useful, you try another query instead of clicking through the results pages.
results pages are mostly kept for nostalgia it seems.
that's a good point about only the first few results mattering.
wolframalpha is computer algebra software turned into a web service which allows for searching some papers, pre-written articles and a bunch of interesting, curated numbers.
wolframalpha provides qualitative answers to "questions about math".
answering the physics problem "whats the speed of light" with "er its fast" is qualitative. saying its 2.99+/- epsilon 10^6 km/s is a quantitative result based on data gathered from some experiment.
answering the search engine problem "whats the speed of light" with "2.99+/- epsilon 10^6 km/s" is not a quantitative result. in a search engine contexts, thats just the most likely correct result . qualitative. just a guess. some computer looked at a bunch of data and decided that one of them was probably right based on some popularity heuristic. its like asking an expert who makes an educated guess and claims its right because seniority.
a quantitative search engine would be capable of answering questions like "how many sites do X".
wolframalpha doesnt do that. its just maple in a browser.
https://www.wolframalpha.com/input/?i=IR+spectrum+of+toluene
http://webbook.nist.gov/cgi/cbook.cgi?ID=C108883&Units=SI&Ty...
First, there are competitors in the space. Some have done an objectively nice job, but they're not blowing up. Second, you don't realize it, but you're not just competing with Google (the search engine) -- you're competing with every product in the Google ecosystem. If you want people to use your new niche search engine, you have to find a value proposition so compelling that people will explicitly choose to use your search, even though every other tool they use is tied back to Google somehow. That's insanely hard.
Finally, but most importantly: programmers are a tiny market, they're terribly picky, conservative, and they don't like to pay for things. If you're going to take on the challenge of beating Google in a vertical, at least make it one that's huge or has huge value. Focusing on programmers means that you'll spend a ton of time trying to optimize for a highly technical, low-value use case that doesn't generalize well.
Solving problems you face daily and are tired of has always been significant driver of innovation. This is an important and satisfying problem to be working on, not all startups have to become the next google or the next unicorn.
No, but you do have to survive. That's hard enough for any company, but it's much harder when you pick a small target market that can't be trivially replicated in other verticals.
So sure, pick a problem that you know well, but be realistic about the achievable market size, and don't start a business that requires huge scale to work (i.e. a search engine), when your target market is tiny.
(Also, as in all things, beware survivorship bias: just because counterexamples to what I'm saying exist, does not mean that the likelihood of creating one is high. Nor does it mean that the counterexamples are doing as well as you think they are.)
A search engine that also has a near monopoly in browsers and phones is a pretty big moat.
Since the bulk of browsing is moving to mobile, it's almost like you have to start with an Android killer and worry about the search engine as a second step.
After our acquisition, we transitioned from small startup where it was super useful to a large enterprise (5500 employees) and I've found that Slack is pretty difficult to scale here beyond engineering/product management.
Slack is strong with helping small scrum teams collaborate, or fielding employee questions in town hall meetings. But, then there's #channel overload which compounds the distraction problem and feels just as bad as email with hundreds of unread messages. And, we have a lot of teams in completely different time zones (India), so the synchronous stream of communication doesn't work as well because it's difficult to track conversations. It's also not inclusive by nature; I have to invite external people to participate, which is a security concern at my company, so it's disabled and used for internal purposes only.
All that said, I love Slack but it is what it is. Asynchronous email is still very valuable to me, much more than Slack.
It was really important for us, because important decisions were being agreed in the midst of a Skype chat, which was only in your history on one machine and not easily searchable.
We also had Google Hangout chats/conversations which are searchable. However this caused further confusion about having to search via email/Skype/Hangouts to find out what was said. Also the conversation from one Hangout was separated from the other - where as we could replace that with one slack channel from all the Hangouts that we had.
Skype for Business now does something similar in that it also keeps you chat history in Outlook.
However I really like the clean separation of emails vs Slack chats. Slack is completely independent of my email so I'm not dependent on Gmail or Outlook.
For me it means that email becomes reduced but I regard it as a channel that is reserved for more important communication, i.e. I'll email something if I want it to be on record as having been written. All the chat fluff gets dumped in Slack.
IRC also has the added benefit of being scalable and free. And everyone has their favorite client already that is more customizable than any Slack client. And you can own your own data/conversations, as an organization. I have never really been comfortable with the notion that Slack can poke around in any channel's history regardless of how secret those conversations are thought to be.
I also can't think of any startup in the 5 years since that has really taken off by pursuing one of these ideas. (Unless you stretch the definitions enough to consider Snapchat a replacement for email.)
So you're right, not many "startups" in here, with the exception of Slack and DDG. I think it's more common for an existing and mature company to take over these spaces when the time is right.
Better for someone to find a better way to organize the world's internet data or links or find a way for me to get more out of it. It's one thing to find me the most relevant website but maybe finding a way for me to do more with that will get me to leave Google.
And if you don't have any ideas then you shouldn't be an entrepreneur because then you're just in it for what, the money? Entrepreneurship is about making something that you MUST get out into the world because you are so driven to do so.
Being an entrepreneur without an idea building what Paul Graham tells you to means you should have a job, not a company.
I'm no expert, but I'd think some of this is already being done, no? However, there's more value to a few companies to have walled-garden messaging platforms vs 100% open and interoperable, and that's not a problem solved by tags.
I really wish the moat weren't so big on this one. The path Google is taking pretty clearly ends with no merit driven organic search results.
But I'll agree that investing in computational power is worthwhile. I say that because of all of this year's news in quantum computing, and because the amount of energy required to represent and flip a single bit should be infinitesimal. So I don't think that it's been shown in principle that achieving higher speeds requires higher (at runtime anyway).
Honestly, it should be NATO initiative. It would be worth a lot more than landing on the moon.
What struck me the most from the essay is the section regarding universities. I don't think universities should be replaced. I think that they should go back to what they originally were, centuries ago, but enhanced by technology, better understanding of the learning process and sensitivity to each learner's personal development.
That means students who are there because they want to learn, not just as a career stepping stone or for rites of passage, and who are free to follow their interests flexibly.
Tutors who are passionate about their subjects, and want to share and who care about the flock they've taken in.
A better peer-learning system; to openly share ideas, adjust and enrich own thoughts and build projects together.
A dynamic curriculum; solid trees of knowledge that are open to wide-ranging thoughts and student needs and whims.
Taking advantage of current technologies, blending them in the curriculum for a better learning experience.
Greater sensitivity to each student's circumstances, who may change their mind to pursue another interest, who may decide to take a break from learning, who may be suffering from crises, ranging from personal to existential.
...
Universities are supposed to be where you can learn freely, and possibly for the undergraduate, to prepare him/her for the real world.
Yet your university experience is measured by what grade you graduate with. No doubt students need assessments - but it should be feedback for further development. Why can't we see the university as an incubator of personal development, where learning is a naturally iterative process?
But it seems that in the "real world", all this seems to be wishful thinking. Universities have to make money! It's not just for commercial reasons, but research needs funding, whole populations need educating (and get jobs.) And recruiters are not that interested in personal journeys, because that's not what companies want - they want newly minted cogs to add to their great machines.
So yes, it is frightening to think about it - it's very messy! In my view, the core issue with universities is that they are not adapting fast enough, and has become too bloated and possibly too commercialised that they have lost perspective of their original mission. So rather than build an entirely new system to replace universities, wouldn't it be more feasible for existing universities to be bolder and make the changes themselves?
The other approach is to provide an alternative learning experience to universities (rather than replacing them.) This is an idea that has been knocking around in my head for a while: how about a public learning centre using MOOCs as the main resource? This will democratise MOOCs even more; you will have richer peer support, greater motivation to complete the course, and local subject matter experts can volunteer as 'mentors' to support the group though with no actual obligation to pass anyone.
Libraries. Libraries can do this.
I believe that his essay here, and nearly all VCs and media pundits, make the same, horrible mistake about planning successful startups:
(1) We have to realize that successful startups are exceptional, one in a million or so. So, for the planning, we are likely planning something exceptional.
(2) A successful startup is usually one of a kind, unique, quite different from the past. E.g., HP, Seagate, Intel, Microsoft, QUALCOMM, Apple, Cisco, Google, Facebook, Twitter, Applied Materials, Amazon, etc.
So, from (1) and (2), in planning a successful startup, we are looking for something both exceptional and new. So, for our plan, we will have to dig enough to be exceptional (the common, ordinary need not apply) and also new (no sense in looking at and trying to build incrementally on the closely related past).
E.g., for the A-bomb, don't go to a bunch of experts in WWI field artillery. Instead, need something both exceptional and new.
E.g., many VCs claim to have "deep domain knowledge", but that has to mean that they spent a lot of time looking at poor business plans and what is in the market now -- not both exceptional and new.
So, (1)-(2) is the horrible mistake.
But, wait, there's more:
(3) Given a candidate startup plan, how to heck to evaluate it accurately? All concerned very much need to know.
Well, good planning of projects both exceptional and new really is possible: Examples include Hoover Dam, the Empire State Building, the Channel Tunnel, Ike's Overlord (invasion of Normandy), the Manhattan Project, turbojet engines, the H-bomb, the US SSBNs, the Boeing B-52 and 707, the engines for the SR-71, the SR-71 in total, GPS, Intel's step to 14 nm, Microsoft's Windows NT, RSA public key encryption, and much more.
So, it is possible to evaluate projects; a project that passes a good evaluation becomes low risk, indeed, for both the technical risk and the market and economic risk.
Then, on (3) PG's essay and VC work in general is short on how to evaluate projects.
My startup is one of the ones on PG's list. I don't regard the project as risky.
Okay, let's see how this goes: The project is a new Web site. The site is totally safe for work and squeaky clean and respectable. The site has only two main Web pages, and both are just dirt simple, so simple they will work fine on essentially any device with a Web browser up to date as of about 10 years ago -- mobile to high end workstation. The project is to solve a pressing problem so far solved at best poorly and shared by essentially every Internet user in the world, especially in the more advanced countries. The project seems to deserve on average at least 30 minutes of eyeball time a week for each of its users -- that's a lot of eyeball hours per year. The site is ad supported with some quite good, new opportunities for good ad targeting.
So, sure, from the above, do some simple, back of the envelope estimation arithmetic and come up with the most valuable company in the world so far -- sure, > $1 T.
Now do we need to look at this "obliquely"? Not really. Once explained, it's all plenty clear and obvious.
Do we need to start on just a small part of the whole problem? Not really.
Does such a big goal have to be a big project? Not really -- I'm a solo founder.
Is the project risky? Now, not really; for me, never was.
Is the project difficult to evaluate? Not really -- I had to evaluate the project and would not have proceeded unless I saw a solid project. But so far I'm the only person who has done any even a little serious evaluation of the project.
Is the project a huge amount of work? Nope, not really. To date it's 24,000 Visual Basic .NET Framework 4.0 programming language statements in 100,000 lines of typing I entered into my favorite text editor KEdit on Windows XP SP3. Nope, didn't use an integrated development environment -- for me learning to use one of those was harder than just typing in the code!
As of now, all the code is as I originally designed it and appears to run correctly. Debugging was no problem. The code is surprisingly fast and efficient (at one point, converting to a solid state disk drive would make the code fast beyond belief). The code is nicely documented. No refactoring needed. The code is fine.
I want to do some more testing, tweak a few places, add some data, and then go live. The code as it is is ready for fairly serious production, and the architecture is easily scalable to serve the world.
So, how can my project be so exceptional? And new? Well, it's new because I thought of it, and on first glance nearly no one would believe that the problem could have a good solution. Well, without some good original research in applied math (the users will never be aware of) with some advanced prerequisites, there is no good solution. So, without looking into the math, and nearly no one in business can, there is no way to believe in the project; so, I'm the only person who does believe in the project.
Nothing "frightening" about it!
So, my view is that the essay is poor on (1)-(3) and, with that poor foundation, makes lots of mistakes, e.g., that big results need a big project.
For an example, I used some of the attributes of my startup.
For my startup, I intend to announce a beta test here at HN.
For any equity investors, I totally gave up: They make the mistakes (1)-(3) which means that they don't like my project and I don't want to report to a BoD with them on it -- on my BoD, after my first presentation of the ad targeting part of the project, they'd all upchuck, soil their clothes, furniture, and carpet on their way, screaming in agony and outrage and terrified of their fiduciary whatever, to the restrooms. After they recovered they'd meet at the closest bar and vote to disband my company. Why? My presentation would have some math in it, math they didn't understand, and listening to math one does not understand for more than five minutes at a time is worse than a root canal procedure. In all their careers they've never seen any role for math. E.g., for ad targeting, they'd be thinking in terms of demographics, gen X, Chablis and Brie set, McMansion set, etc., and none of that would be in my targeting math.
The equity investors didn't want me when I started writing the code; now I don't want them; and as soon as my project gets any reasonable revenue forever I won't want them, their check, their BoD, or anything else. My project is to be exceptional; equity investors are not exceptional.
Sometimes ambition is great enough even when finances are inadequate, where just moving a little bit in the ultimate direction (like shipping hardware) is materially out of reach for such an extended period, startup and growth potential can never be realized without the most exceptional source of funds to provide leverage.
This pushes the time horizon further out as it would naturally take much longer to connect with such an elusive benefactor on terms that can truly maximize the performance of the combined resources.
So it seems like a good idea to have ideas that can stand the test of time, the more ambitious you are.
Great advice from pg about tactically limiting the depth to which your full ambition is displayed day-to-day so that rather than pie-in-the-sky being constantly pursued, stakeholder interactions can be punctuated by achievable milestones they can visualize being reached.
In a capitalist arrangement it can come back to the visionary exceptionalness of the capitalist as the limiting factor as to when or if you can bring your full potential squarely within sight.
Make a new spam free telephone system on the ip network
I've been expecting and actively looking for this for more than a decade. I have never found anyone discussing this idea, even less attempt to implement it. People really can't imagine what the post-app world will look like. Even people here on Hacker News. I don't expect most VCs to appreciate it either. I'm on my own, and I hope people will realize how important this is when they'll see it.
The key to understand this is to realize that it's not just about building another app. It's not just about building another OS. It's not just about building another programming language. We need to completely rethink the way we think and communicate. We need to invent a new communication paradigm, a new medium for thinking.
The goal should be to eliminate the gap between the thought of something and the existence of something. I should be able to think about X, and X should then exist. That's the gist of it.
Obviously, it's unreasonable to expect that gap to disappear within the next 10 years. However, I believe that 10 years is enough to reduce that gap by more than it was reduced in the past 1000 years.
Google asked themselves what email would look like if it was invented in 2009. I'm asking what language would look like if it was invented today. Who else attempted anything like this in recent history?
Now back to something more concrete you can all relate to. How many apps do you have on your phone? How many online accounts have you created in your life? How many different messaging apps are you using? Most people are very adaptable, and don't realize that the answer is "too many". We've seen many attempts to improve this situation, with things like the semantic web, OAuth, WeChat, personal assistants, ethereum, Vurb, bots, Google Lens, etc. That's not sufficient. We need something much more ambitious. We need real disruption.
I intend to make 50% of existing jobs obsolete within the next 10 years. I'll need some help. Who's in?
"Single point of failure," "Separation of concerns," and "Don't put all your eggs in one basket" are the phrases that come to mind. Risk mitigation is a concern for many of us, and what you describe sounds Orwellian to me.
Have you checked out Urbit? I believe it's a decentralized/federalized system instead of the centralized on you describe.
But, I'd be interested in seeing your prototype of The Borg :)