1. Hundreds of pitch meetings with VCs. Plus they are pretty open about their expectations.
2. Fire you. Sue you for fraud. Talk badly about you in the media. Not saying they would, but there is a lot a deep pocketed VC could do to make your life miserable. It's generally a bad idea to lie to your investors. :)
In practice, the main difference is the pay / relative "rank" within a company. Employer's perspective: Want to hire a software engineer but don't have the budget? Create a new role, call it something else, now the pay difference is justified.
Context - I'm a software engineer by title but my job is application development. Which is hard, mind you! It looks like simple coding but it involves prioritizing, making smart decisions, and making all stakeholders happy.
Dating apps/sites that employ the strategy of "get as much data as possible and match intelligently" have failed to compete for users against Tinder, the dumbest, simplest dating app that hardly does any intelligent matching at all. This implies that it's not the sophistication of a dating app that matters, it's the simplicity and ability to attract a userbase that in turn attracts a larger userbase etc in a virtuous cycle.
Tinder is doing a lot more intelligent matching than you think. The CTO of OKCupid (which is owned by the same parent company) was on a podcast a few weeks ago where he briefly discussed some of the algorithmic challenges of Tinder. They include:
- it may not seem like it, but Tinder does try to pick people it thinks are most likely to produce a match, based on-- to be brutally honest-- your attractiveness (as determined by who has swiped for you) and the other person's
- it then orders them approximately from most likely to least likely, but it has to know how often to start over, because eventually as you go down the list you'll eventually start running into people that you're less likely to swipe right on than the people you already passed up
- carefully arranging things so that "super likes" have a reasonable chance of producing a match while also not inundating very attractive people with nothing but super likes from people they will never respond to (which would drive them off the service)
It's a tough challenge and Tinder is not at all a "dumb, simple" dating app.
Maybe he missed out on some of the other important things Tinder does on the "addiction" side of things.
When the app first came out (or possibly when you first start using it), the first person is/was always a match. Then they got clever about hiding matches deeper in the swipes.
This triggers addictive behaviour because the "match" reward is inconsistent. Though, clearly, before you start swiping, Tinder knows which matches it is going to present to you. It's a delicate balance to judge how long you will spend swiping, when to show you your match and how to make sure you swipe right a few times to keep the funnel loaded.
This bolsters the point that the future of online dating is not a computer estimating compatibility with creepy accuracy, it's someone engineering an app to be addicting to get the largest userbase.
Understood. But they have so far avoided asking the user for structured data that they could use to vastly improve their matches. Clearly it took a back-seat to usability.
Perhaps I shouldn't have used the word "dumb" - I do not mean to imply the Tinder strategy or algorithms are stupid. Just meant that instead of prioritizing matching to the best of their ability, they prioritized usability.
Also, for dating sites in particular, doing an excellent job of matching can be bad for business. Successful matches, especially for users seeking long-term monogamous relationships, mean they stop paying for your service. This is most extreme for the services that are practically marriage brokers, and that's where y ou see the "lots of data" approach most often.
At the other extreme, apps like Grindr and to a slightly lesser degree Tinder aren't hurt by doing "too good" a job, since more of their users don't stop looking because they found one good match.
I'd argue that the market of 'people who are dating' isn't going to be dramatically affected by any particular company, and probably not even by online dating in general. They would have to be stupendously effective at creating long-term monogamous relationships to really move the needle there.
Marriage rates have been declining for decades, and serial monogamy seems to be growing (can't find good quantification of that). Successful dating services don't necessarily cannibalize their market.
My intuition is that providing an effective and enjoyable experience is going to provide far greater returns by gaining market share in a relatively inelastic market.
> Successful dating services don't necessarily cannibalize their market.
GP is clearly not suggesting that a dating service can be so effective that it depletes the market of single people. They're suggesting that a successful dating service won't retain customers, insofar as customers who find their life partner no longer need dates.
Possible that you are not giving enough credit to Tinder's algorithm for sorting the profiles that you're presented with? There's a lot of information buried in how you swipe and message and edit.
I was so on-board at the beginning - yes, html and css skills are undervalued.
And then the essay tried to generalize the lack of respect for HTML/CSS specialists as an example of lack of respect for specialists compared to full-stack devs. This is where I think the essay took a wrong turn.
The technical community has tremendous respect for backend or math specialists. Now more than ever, ML and data science is cool. High performance is cool. Security and cryptography are cool. Etc.
But HTML and CSS? There is a general attitude that they're "easy" and javascript is "harder". But writing good HTML and CSS should be considered one of the highest art forms. Good HTML and CSS makes the end user go "wow, this is nice and clean" and makes the javascript developer go "wow, this is nice and clean". Achieving both of these is so fricking hard!
Only true as of language. If you're doing websites you get to learn idiosyncrasies of browsers, DOM manipulation, asynchronous design, performance considerations, etc. That on top of how HTML and CSS is laid out and converted to DOM.
Which is as hard if not harder.
Javascript is not easy to learn. Many language quirks, and doing anything fancy with asynchronous code will require an understanding of closures, functional programming, and perhaps promises.
For 1 and 3, it's the process that matters more than the tool. One process I found effective is maintaining a spreadsheet of tasks and a spreadsheet of larger milestones, updated in periodic meetings.
For the task spreadsheet, you can have a twice-a-week stand up where everyone goes around and updates their own tasks on a google sheet in turn.
For the milestone spreadsheet, a period between one month and one quarter works well.
Although meetings are often eschewed by software engineers, they really do help to keep projects on track.
Also appoint someone to be in charge of tracking the "health" of various milestones, red = needs course correction, orange = at risk, green = on track, etc.
Decentralization and auditing. If you put some code on AWS there are two potential issues.
Firstly you don't know if AWS has actually run your code. They could potentially replace your code with their own modified version which does something malicious (like add a bias). For example in a voting system, code could be modified to ignore X percentage of unfavourable votes, or change unfavourable votes to a more favourable candidate.
Secondly I don't know what your code on AWS does. This isn't great for financial and voting systems that should be audit able.
With Ethereum you can view the code for a contract to validate it does not "cheat". You can also trust the contract was executed correctly because the miners have validated it and added it to the block chain.
well, not specifically ethereum but stuff like siacoin provide cheaper options than AWS and while I personally cannot attest to the performance, they seem to be interesting alternatives.[1]
In the case of siacoin, at face value it seems to be significantly cheaper than AWS or Google Cloud for providing bulk cloud storage.
Even more exciting for me is IPFS and Filecoin. Seems like this pairing could allow for a truly decentralized Internet. Ycombinator had one of the founders / lead programmers on the podcast in the last few weeks and I highly recommend it!
siacoin is relatively slow, it took me ~2 minutes to upload 40mb of stuff. But really it's meant to be cold storage and at $2/mo/TB it isn't that bad. The speeds will also increase with adoption.
I would imagine the startup life skews young in general because older age tends to mean more responsibilities, and at some point if you haven't been successful with a startup you'll tend towards a more traditional position for those reasons? "Startup Veteran" is kind of an amusing phrase.
I disagree. I'd say older people have more networks, more experience and more resources, hence will find less comparative benefit in giving away precious equity to YC (etc) than young people who need all the above.
It could be said that YC and other startup incubators have done a stellar marketing job in taking ~6% equity off organisations by turning the tables on funding and making companies "apply" to give their equity away. From YC's perspective they have nothing to lose and everything to gain from this arrangement, versus the founders who are in pretty much the opposite position.
Youre confusing concepts. You can skip an incubator like YC while still intending to not bootstrap. There is also no such thing as a 'YC-style' startup, YC was set up to target startups, so can't have invented them.
YC used to only give a tiny amount of equity that any vaguely successful person could match ($16k?). It filled the gap between starting and series A.
Bootstrapping means avoiding getting a series A.
Basically, YC didn't invent startups, it invented a new, more accessible route to starting one.
When I picture a "Startup Veteran", I picture someone who's worked with one or more companies to a large and successful exit. Someone like that is likely fairly wealthy, and so doing a startup doesn't have the same amount of risk as someone doing it for the first time has.
I would expect that with a fairly even distribution such as ages, the median and mean would be similar. Medians are primarily useful for data sets with large outliers and I doubt there are many 400 year old YC alumni :)
Not sure, but if there are a bunch of 40s years old, this would already start shifting the average upwards compared to the mean even if the data is rather centered in the mid 20s.
