Interesting move. I've just read Peter Thiel's startup notes, and one point he's adamant about is to build your product as a monopoly for a small market (this is in no way new advice, but since I just finished his book his formulation is the most recent one in my head). E.g. Facebook and Harvard students, AirBnB and housing for conferences, etc. If you can't get to mass adoption in a small niche, how are you going to scale it up? Additionally, your product has to be at least an order of magnitude better than what the niche already uses (e.g. Facebook was an order of magnitude better than Harvard's static face book).
I have a hard time seeing what small niche Eve will dominate, and how it's at least 10x better than any existing solution. Looking at the examples they give:
> a scientist doesn’t have to rely on the one person in the lab who knows python
Scientist these days pretty much have to pick up a programming language (Python, R, Matlab, etc.) throughout their undergrad/early grad studies. If you're going to target that niche, not only are you fighting against the huge investment professors and labs have made in existing tools (along with the small libraries/frameworks they built themselves), but you also have to provide an equivalent quality ecosystem for the weird data formats some very specific niches use, for the lab equipment, and so on. If you wanted to deliver a 10x improvement on the existing state of computation in research labs, you'd have to spend a lot of time focusing on exclusively that.
> a child could come up with an idea for a game and build it in a couple of weekends
They already can, using Scratch or the myriad similar tools available which will probably remain better because they were designed with game design in mind.
> your computer can help you organize and plan your wedding/vacation/business.
People already use MS Office, Google Docs, etc. for these tasks. Unless Eve brings a 10x improvement to those tools, how can they ever hope to displace them? My dad has been using the same version of Excel for the past 10 years, I have a hard time imagining people like him switching over.
In his lectures, Peter Thiel spends some time talking about all the solar startups that emerged more or less at the same time in the late 2000s, only to fail miserably. While the buzz and excitement for solar power companies was there, in reality they could not deliver on what startups need to deliver to thrive: a product solving a real problem for a niche market that's at least 10x better than the existing solution in place. I feel like we're in a similar space with all those "learn to code" and "new programming tools" companies that have been steadily emerging.
No doubt, the CS academic and HCI researcher that I am is excited by Eve- I'm sure they've done their research and that it would be a really neat product.
The entrepreneur in me is however extremely doubtful.
Our first bastion is websites, you can think of it as filling in the gap that Microsoft left when they killed VB6. 15 or so years ago, someone could open up VB6 draw out a couple of forms and put together a simple "Line of Business" app themselves without having spent 10 years of their life becoming a professional programmer. The modern VB6 would do the same for websites. :)
> Unless Eve brings a 10x improvement to those tools
We know the bar is pretty universally an order of magnitude and so we've planned accordingly. Our primary target here is selling a super power: you couldn't do this at all before, and now you can. Joe in accounting doesn't have the opportunity to build things to support himself (short of studying programming), but with Eve he could do everything from automate several of his workflows, to building a dashboard, to building dynamic reports for his bosses, to creating internal tools to track and measure things.
There is a fairly clear gap in our ecosystem around these kinds of programs and they represent what I would argue is the vast majority of programs that should be written, but aren't because of the cost.
There's also the opportunity to just make websites an order of magnitude easier to build for programmers and we're getting there. I was able to write an entire editor for this thing in itself in a week for one of our prototypes - there's some real potential there.
> 15 or so years ago, someone could open up VB6 draw out a couple of forms and put together a simple "Line of Business" app themselves without having spent 10 years of their life becoming a professional programmer.
There's plenty of people throwing together simple LOB apps (including, particularly web apps) now -- with a wide variety of different tools -- without spending a decade becoming a professional programmer. And using similar workflows as they did when VB6 was around. Visual designers with minimal code didn't go away when VB6 did, and the place where there are the most competing ways of doing this is on the web.
> Joe in accounting doesn't have the opportunity to build things to support himself (short of studying programming), but with Eve he could do everything from automate several of his workflows, to building a dashboard, to building dynamic reports for his bosses, to creating internal tools to track and measure things.
That's exactly the pitch for lots of things that are on the market now (to pick one among many, without any particular intention to claim its the best in the field, Zoho Creator).
> Our first bastion is websites, you can think of it as filling in the gap that Microsoft left when they killed VB6.
They've created LightSwitch. LightSwitch is also a great example for the usual outcome of using RAD tools: You often come pretty soon to the point where the provided abstractions are lacking some functionality or performance. The customer / user doesn't care how the software is built, it's compared to the greatest there is.
Bricklin was joined by Bob Frankston, and the pair worked on VisiCalc for two months during the winter of 1978–79, forming Software Arts. Bricklin wrote, "[W]ith the years of experience we had at the time we created VisiCalc, we were familiar with many row/column financial programs. In fact, Bob had worked since the 1960s at Interactive Data Corporation, a major timesharing utility that was used for some of them and I was exposed to some at Harvard Business School in one of the classes." Bricklin is referring to the variety of report generators that were in use at that time, including Business Planning Language (BPL) from International Timesharing Corporation (ITS) and Foresight, from Foresight Systems. However, these earlier timesharing programs were not completely interactive, nor did they run on personal computers.
I have a hard time seeing what small niche Eve will dominate, and how it's at least 10x better than any existing solution. Looking at the examples they give:
> a scientist doesn’t have to rely on the one person in the lab who knows python
Scientist these days pretty much have to pick up a programming language (Python, R, Matlab, etc.) throughout their undergrad/early grad studies. If you're going to target that niche, not only are you fighting against the huge investment professors and labs have made in existing tools (along with the small libraries/frameworks they built themselves), but you also have to provide an equivalent quality ecosystem for the weird data formats some very specific niches use, for the lab equipment, and so on. If you wanted to deliver a 10x improvement on the existing state of computation in research labs, you'd have to spend a lot of time focusing on exclusively that.
> a child could come up with an idea for a game and build it in a couple of weekends
They already can, using Scratch or the myriad similar tools available which will probably remain better because they were designed with game design in mind.
> your computer can help you organize and plan your wedding/vacation/business.
People already use MS Office, Google Docs, etc. for these tasks. Unless Eve brings a 10x improvement to those tools, how can they ever hope to displace them? My dad has been using the same version of Excel for the past 10 years, I have a hard time imagining people like him switching over.
In his lectures, Peter Thiel spends some time talking about all the solar startups that emerged more or less at the same time in the late 2000s, only to fail miserably. While the buzz and excitement for solar power companies was there, in reality they could not deliver on what startups need to deliver to thrive: a product solving a real problem for a niche market that's at least 10x better than the existing solution in place. I feel like we're in a similar space with all those "learn to code" and "new programming tools" companies that have been steadily emerging.
No doubt, the CS academic and HCI researcher that I am is excited by Eve- I'm sure they've done their research and that it would be a really neat product. The entrepreneur in me is however extremely doubtful.