Thanks for being here!
This is the Alternative Data Weekly for Friday, April 19, 2024.
Announcement(s):
Excited to let everyone know that I am joining the team at Module Q. See more in this week’s Final Thoughts section below.
Theme that emerged in this week’s email is … GenAI is impacting the investment world in a multitude of ways.
QUOTES
“Bad data strategy isn’t just one that doesn’t deliver, or one that fails for some reason. It is a very common way of getting data wrong. In fact, today, the default way to go about data is the bad data strategy.” Sven Balnojan
News Articles
Podcasts
Cool Charts
Final Thoughts (ModuleQ)
#1 – Doug Hopkins published Transforming Finance: The Impact of Generative AI on Financial Data in the Past Year. March 2024.
My Take: AI is a new tool that is driving change in many areas of life, but finance has proven to be an early mover and, given it is a data-heavy world, finance has the potential to see a lot of benefit.
Doug articulates some of the initial problems of AI in finance (see Doug’s previous article from June 2023). Data privacy and veracity are big issues as finance is reliant on data that is accurate and strictly adheres to compliance requirements.
It will be fascinating to watch as new companies & categories of companies are created to service this market.
#2 – Vin Vashishta’s What Is AI-Driven Growth?. March 2024.
My Take: Everyone will eventually be using AI (even if they don’t know it). The author provides an interesting maturity model framework. Until a business has systems, models, and frameworks that explain complex & uncertain AI product strategy concepts, there’s no way to realize an AI-driven growth trajectory.
Zombie
Legacy
Competitive Maintenance
Competitive Advantage
Innovator
Accelerator
Foundation
BONUS: NIST published NIST Identifies Types of Cyberattacks That Manipulate Behavior of AI Systems. January 2024. ”One major issue is that the data itself may not be trustworthy. Its sources may be websites and interactions with the public. There are many opportunities for bad actors to corrupt this data — both during an AI system’s training period and afterward, while the AI continues to refine its behaviors by interacting with the physical world.“
What else I am reading:
Alex Izydorczyk’s Timeseries, Transformers, and Two Cultures. April 2024.
Ethan Mollick’s On the necessity of a sin. March 2024.
Three Data Point Thursday 12 Memes Of What I Learned About Navigating The Data & AI Space Of Tomorrow. April 2024.
Jason Derise’s Can Europe Continue to Fuel Netflix's Global Growth? An Interview with Fable Data April 2024.
CiaoLink’s Report: Data Industry Job Changers March 2024
Wendy Turner-Williams’ Unleashing the Power of Data with Trusted AI. April 2024.
Eagle Alpha’s Consumer Transaction Data: Unveiling Trends and Spending Patterns. April 2024.
I came across a few EY articles … I am curious how changes in payment tech will impact data used for investing … particularly the “tables stakes” credit card data.
How the rise of PayTech is reshaping the payments landscape. October 2022.
How Gen Z’s preference for digital is changing the payments landscape. January 2024.
The related Alana Semuels Time article: Why We’re Spending So Much Money. March 2024. (hint: frictionless payments)
Source: The Analytics Engineering podcast published AI’s Impact in the World of Structured Data Analytics (w/ Juan Sequeda). March 2024.
My Take: Great perspective on how we got where we are today regarding data analytics. Zoom out. Good idea session about how things are changing in this world of data.
Juan is bullish on meta data, semantics knowledge, data catalogs … this will be the heart of AI for enterprise (minute 42:00).
Of most interest to me was the discussion of the importance of meta data. Meta data has been left on the side … meta data is essential as we move from data first world to a knowledge first world (Minute 22:30).
One interesting idea: when request a meeting … your prep is (will be) talking to their “AI specific domain” to which they have uploaded relevant information. This will make the meeting much more productive. I translate this as the amount of information available to you prior to the meeting will be staggering, the real key will be how to best make use of it…AI can help deliver only that data that is relevant and important.
Another interesting point reinforces that you get what you incentivize. Data.World bonuses employees on customer adoption. Result: >2M users on their open data catalog.
A data catalog should allow you to catalog more than just the data … should be broader than that, and ultimately give you context of the search. The term catalog feels to passive … there is so much more.
What is the one thing Juan is most hopeful 5 years from now? Treat “Knowledge engineering” work as a first-class citizen … focus on “investing in knowledge”.
Other notes:
Three things to do with data: Move Data, Store & Compute Data, Use Data
Academic benchmarks separated from the commercial world.
Trust: Accuracy, Explain-ability, Governance.
Juan also hosts the Catalogs & Cocktails podcast.
Highlights (48-minute run time):
Minute 01:30 – interview begins
Minute 02:30 – Juan’s background (semantic web); interesting combi of academia & industry
Minute 11:30 – the advances of analytic tools not being taken advantage of today
Minute 13:00 – data.world overview (enterprise data catalog)
Minute 16:45 – catalogs as a category; things get buzzword-y
Minute 21:00 – the life of a data analyst in 2030
Minute 24:45 – conversation gets more into AI
Minute 29:30 – Juan’s vision for 5 years from now
Minute 34:30 – observability vs analysis
Minute 40:00 – GOFAI = Good Old-Fashioned AI
Minute 45:00 – analogize a typical meeting to LLM
Source: A16Z’s Sarah Wang & Shangda Xu published 16 Changes to the Way Enterprises Are Building and Buying Generative AI. March 2024.
Budgets for generative AI are skyrocketing.
Leaders are starting to reallocate AI investments to recurring software budget lines.
Measuring ROI is still an art and a science.
Implementing and scaling generative AI requires the right technical talent, which currently isn’t in-house for many enterprises.
A multi-model future.
Open source is booming.
While cost factored into open source appeal, it ranked below control and customization as key selection criteria.
Desire for control stems from sensitive use cases and enterprise data security concerns.
Leaders generally customize models through fine-tuning instead of building models from scratch.
Cloud is still highly influential in model purchasing decisions.
Customers still care about early-to-market features.
That said, most enterprises think model performance is converging.
Optimizing for optionality.
Enterprises are building, not buying, apps—for now.
Enterprises are excited about internal use cases but remain more cautious about external ones.
We believe total spend on model APIs and fine-tuning will grow to over $5B run-rate by the end of 2024, and enterprise spend will make up a significant part of that opportunity.
Source: ModuleQ
I am excited to join the team at ModuleQ.
ModuleQ is an Unprompted AI company focused on delivering the right data, to the right person, at the right time, directly inside their workflow.
As someone who has spent their entire career in research & data, I am all too familiar with the issue of ensuring frontline knowledge workers like PMs, analysts, bankers, & wealth managers, have access to not only relevant data, but the best data, at the right time. While many of us have worked on the data & insights problem, few of us have tackled the delivery problem.
Enter ModuleQ.
ModuleQ has solved the last mile delivery problem by connecting to all of your data & delivering it to you in a hyper-personalized experience in your collaboration tool of choice (MS Teams, etc). See video demonstration.
As I have written these ADW’s over the past few years, it has become clear how revolutionary AI tools can become. The proactive nature of #unpromptedAI simply makes you life better. You’ll be interfacing with AI without knowing you are interfacing with AI … it will just be how things work.
You’ll be more productive, have more opportunities, and be connected in a much more relevant and personalized manner.
You can expect the ADW weekly cadence to continue. Data & AI are inseparable concepts.
I look forward to sharing more about my experience at ModuleQ in coming weeks/months.
Can’t wait!
New Email: john.farrall@moduleq.com