Hello,
Wanted the board to be aware of this new interview with Upstart CEO, Dave Girouard:
https://www.buzzsprout.com/1697077/8693359-investing-in-ai-e…
My takeaways:
On their edge:
-Enormous data set currently limited only by computing power
-Their AI is made up of a series of models blended together
-Highly specialized personal lending data set
On banks as competitors (and banks having more “data” than Upstart):
-Banks can’t just press a button and apply their data into productive data sets/training data
-Upstart has a highly specialized “machine learning team” composed of PhDs in Math/Stats- “If I had to guess I would say there might be 100 software engineers for every 1 machine learning engineer who can build these types of models”
-Very risky for banks to experiment with their data. Upstart was able to take on that risk as a VC-backed start-up
On why they lend money directly (2% of revs):
-“R&D purposes ONLY”
On the future of the company:
-Personal lending can expand for at least 5 years before reaching a TAM ceiling
-Expanding outside of personal lending: The sky’s the limit! (my words, lol!)
After listening it dawned on me how huge the potential of this company is. Added to my position this morning.
*I would also like to thank the board for recent discussions. My take: the future is unknowable; focus on present performance and follow the numbers.
-SF (Long Upstart)
58 Likes
I found the podcast linked in the prior post extremely valuable. 
Some of my notes that were left out of the prior post:
Paraphrased from the CEO of Upstart-
Upstart vs Big Banks is like Tesla vs Ford
Tesla is able to collect usable data to inform the Self-driving AI models from all their cars on the road.
Ford has Millions more cars that have been on the roads for decades longer. Ford can not use any of that data to inform the development of any AI models.
The Moat for Upstart is their proprietary algorithms, developed by an unparalleled Machine Learning Team, that now ingest an Exponentially growing Data set, now so large that the amount of compute on AWS limits Upstart in their ability to further develop their AI models.
Regarding Big Banks not being able to take on the Bias Risk inherent in using customer Data to develop AI models-
CEO of Upstart, Girourd, said, I paraphrase “It’s impossible for me to see how , even if Big Banks were able to hire enough ML Engineers, they would be able to take these risks inherent in developing any type of successful AI Model that a Venture Capitalist funded Start up could take- like would it be able to follow decades of Fair Lending Laws, avoiding Biases, and be successful at increasing the numbers of loans given and decreasing default rates? I just don’t see them as being able to take those risks.”.
And…I continue to paraphrase-
“Personal Loans and Automotive Lending TAM would keep Upstart growing at current rates for at least 5 years; but, we’re planting flags in all areas of lending and if successful will be the biggest Fintech company in the world”.
Yes, after listing to this podcast I also gained in my confidence level and added to my position taking it up to 8.5%
Best
Jason
27 Likes
Dave Girouard said around 11:55:
“The algorithms that we have built are far more important than the data, because there will be a time where we will jump over a fence and all of a sudden be creating a new product for a new market where the data is completely new and different. Yet what we will start with is our algorithms that can make sense of it.”
It would seem Dave Girouard is implying that Upstart is not simply caged in to just loans and the banking industry, and that we might see them jump some fences in the future. Certainly, if they are jumping fences the TAM will grow with each fence they jump. I wonder which fences they might jump?
W
13 Likes
They currently use their AI & ML for marketing and customer acquisition so it would seem likely they would create “products” of these capabilities and then look to sell them to any company (not just financial companies). That would massively increase the TAM. I do see them as a technology company who have a “module” for loans and I fully expect them to release many more “modules” over time. Very exciting company.
6 Likes
It would seem Dave Girouard is implying that Upstart is not simply caged in to just loans and the banking industry, and that we might see them jump some fences in the future. Certainly, if they are jumping fences the TAM will grow with each fence they jump. I wonder which fences they might jump?
Healthcare is ripe for AI deployment. As a former healthcare professional Saul might hate me for this but whilst I would prefer for Drs to diagnose and perform procedures, AI/programatic prescribing could be a game changer.
Education is also ripe. Personalised AI powered eLearning could be a game changer.
Any decision making process where quality can be improved or capacity raised where supply cannot meet demand.
Ant
3 Likes
It would seem Dave Girouard is implying that Upstart is not simply caged in to just loans and the banking industry, and that we might see them jump some fences in the future. Certainly, if they are jumping fences the TAM will grow with each fence they jump. I wonder which fences they might jump?
I wonder if they could revolutionize the credit card industry? A lot of the customer reviews on Credit Karma are from people who refinanced their credit card balances using Upstart. Why couldn’t Upstart just come out with their own credit card (or license their tech to an existing credit card company) that has much better interest rates?
I don’t know much about the credit card business, but it seems like having people accumulate a credit card balance only to refinance with Upstart is an inefficiency that could be eliminated. And to see where Upstart might be in the future, Visa has almost a $500 billion market cap whereas Upstart’s is what? Less than $20 billion?
Upstart definitely has potential.
2 Likes