Database Reference
In-Depth Information
who says that in ten years we're going to solve general artificial intelligence.
I don't think a lot of people in that class recognized the degree to which the
things that Vicarious was saying are the things people have been saying for
the last twenty years, as it's very easy to overhype and overpromise on these
sorts of technologies. Then you had us, Prior Knowledge, who sat right in the
middle of the spectrum.
Personally, I think that the best thing that happened in the class, which I'm
not sure it ever made it into the notes, was that Peter asked, “Do you ever
worry about your technology being used for evil?” The people at Palantir said,
“Well, yes, in fact, we have an entire legal staff dedicated to making sure that
that doesn't happen.” I think that's the best spin I've ever heard on a compli-
ance department. Having a bunch of lawyers because you like government
contracts is very different from saying no, we employ John Connor.
Having Peter teach the startup class was great because Peter's extremely
smart and very willing to make long-term technology bets. He's excelled
in investing in science and state-of-the-art engineering as well as pure-on
commercial consumer investments, like Facebook. There aren't very many
VCs who make those kinds of bets, and Founders Fund has always been very
much willing to take the risk here for deeper science plays. Of course, I'm
obviously very grateful for that. Finally, it was fascinating to watch the class
reach a certain degree of infamy due to Blake Masters taking extensive notes
and posting them online.
Gutierrez:
Could you turn your neuroscience research into a startup?
Jonas:
I wish it were possible to make a billion dollars developing technology
that would make neuroscience better. I wish it were possible to make more
money building tools for science because you could hire an engineering
staff and a support staff to help speed up progress. However, it's just not
sustainable. The reality of the current situation is that science tool companies
are a shitty business. If there's a very clear and useful clinical application for
a product, then often it's more viable. Illumina, which makes systems that
analyze genetic variations and biological functions, is an example of this type
of company.
Unfortunately, at this time, I don't think there's a way to build a startup around
neuroscience tools. Even if I could accomplish something as dramatic as build-
ing a tool to record from a billion neurons, which I think everyone recog-
nizes would be one of the biggest breakthroughs in neuroscience in the past
50 years, the total addressable market for this tool would be around 20 labs,
each of which has a couple of million dollars in funding a year from the NIH
if they're lucky. It's not a market that can sustain large companies. I wish it
was—but no, alas.
