Looking Forward (Heavy WIP) 👉
I'm consistently inspired when I meet teams that are working on world changing ideas.
I do my best to befriend and support the dreamers and builders. Don't observe. Build.
Here's a list of future technologies and trends I believe in. And the companies working on them.
Hit my line if you're building something cool. Would love to chat about it, provide you a signal boost, etc.
luke [at] spawner [dot] ai
First published: 11/2/2020 || Last updated: 11/4/2020
Access and Decentralized Finance
Summary: We're seeing financial access accelerate at a neck breaking pace. We're opening up financial systems to more people than ever with apps like Robinhood and Public. And in crypto land we're completely reimagining complex instruments as trustless protocols with completely new systems of value and trust. Crypto is enabling all sorts of interesting open finance use-cases. Bitcoin is neat in principle, but the access to previously closed systems is what's most exciting (to me). Complex financial products typically have been reserved to sophisticated market participants. Now these are becoming open and decentralized. Complex economic transactions are happening between otherwise unsophisticated market participants. The average user is still incredibly tech savvy, but hopefully we can start to build useful lending, borrowing, and transacting platforms for everyday users in the next few years. Until then, I'll be farming yields and losing small bets to unaudited contracts, failed governance, and an untested environment full of fresh game theory and new types of transactions.
Companies: See my crypto investing section
AI in Manufacturing
Summary: I'm hyper bullish on AI in manufacturing. So much so that I went and got a book deal to write about it. There are a number of companies innovating in this space, but the transformation has only begun. Many of the most interesting problems to solve with AI are coming forward as we unlock vision, sound, and other mediums of data transfer. It can start in lower stakes manufacturing environments and eventually permeate across industries. We of course don't have SaaS margins, but we have massive industries with global manufacturing lines and repeatable models and deployments.
Cybersecurity as a Service (CSaaS)
Summary: Most companies struggle with cybersecurity. Only the top companies have the excess cash to spend on large security teams. And those who do have the money end up spending massively on infrastructure and talent to maintain a decent security posture. Even then, it's many times lacking. Many companies that do business online or have a large share of their workforce as knowledge workers treat cybersecurity as a third-class citizen. It's not affordable for most. And yet the stakes are extremely high.
In startupland cybersecurity is notoriously overlooked. Roll the dice on Product Hunt and every third product will have some exploit staring you in the face. This is the story across much of early stage technology companies. I've sent a handful of security issues to teams in the past, and I'm by no means educated on anything security related.
CSaaS is a natural progression as companies outsource their Cybersecurity to automation, external teams, etc.
The most interesting innovation in this space is bolt-on security solutions. Sqreen is my favorite in this space, and there are plenty others.
Collaborative Data Science
Summary: Data Science has been making its way into most companies over the last decade with the massive hype. What's even more interesting at this point is a trend towards empowering citizen developers across corporations. Knowledge workers everywhere are realizing they can also find benefits from understanding the business's data inside their function. Imagine the financial analyst getting access to Tableau. The social media manager getting empowered with Looker. Or imagine the R&D and product designers getting easy access to customer feedback. It's really resonating with lots of folks.
What's interesting now is how companies can take the next step in this evolution. I really believe in collaborative tools for this next step in the evolution. I think many are still a little too complex to be useful. Deepnote, Count, and others are building really great collaborative tools. But there's still a massive whitespace for collaboration without needing to write a bunch of code. No-code Data Science is inevitable.
In the future, teams will embed data professionals who will help them take the next step. And collaborative tools are the natural progression to get there.
Machine Learning Infrastructure
Summary: This is still incredibly lacking. Most of my Machine Learning friends are still building loads of infrastructure around their solutions to serve at scale. There are countless tools (bug-ridden) from countless builders. Lots of more sophisticated teams are open-sourcing great stuff. But ultimately there's no clear playbook. Expect infrastructure to consolidate and some ML infrastructure teams to really hammer down making deployments just a few clicks. Massive DevOps hurdles still stand in the way of many deployments, and we can expect these to be automated. Plenty of companies working on this stuff.
Retail Trading Platforms & Tools
Summary: Retail trading has been growing year over year for some time. Covid accelerated this trend and Robinhood has been the natural profiteer. Other brokerages have also massively benefitted. TD and other brokerages continue to grow. What's interesting is the alarming lack of tooling available to active retail traders. Active traders have a high willingness to pay so products to serve these users are inevitable. New products are built every year but they're stale, samey, and lack any sort of real spark or gamification. Expecting a flood of products to address the massive opportunity created by the ill-informed retail traders of yesterday, today, and tomorrow.
Summary: I'm bullish on the companies building practical applications with RL. Robotics is ripe for this. I've talked to a number of teams that have unlocked fascinating use-cases with RL, especially in shipping with sorting and handling, in hard industries like chemicals and construction as well. Lots of interesting stuff. The issue here is the sheer number of teams with seemingly little advantage over the others. The massive industrial companies like Siemens and Rockwell are pretty entrenched, and it's easy for them to convince long-time customers to use their advances in AI over some untrusted newcomer. Startups will need truly advantaged tech and teams to have a chance in these arena (i.e. Covariant).
This list is ever-evolving. Thanks for reading.
You can email me at: luke [at] spawner [dot] ai