Alternative Data Weekly #222

January 24, 2025

Jan 24, 2025

Thanks for being here!

The theme that emerged in this week’s email is … simple trumps complex.

QUOTES

“Success in the LLM space isn't about building the most sophisticated system. It's about building the right system for your needs. Start with simple prompts, optimize them with comprehensive evaluation, and add multi-step agentic systems only when simpler solutions fall short.” Erik Schluntz and Barry Zhang of Anthropic Building Effective Agents

News Articles
Podcasts
Cool Charts
Final Thoughts (K.I.S.S.)

#1 – Conner Folley published on World of DaaS AI is a $200 billion market for data providers. January 2025.

My Take: I love the analogy to the cloud & SaaS markets. The data market is at an inflection point. This is what is needed to unlock the value of data:
- Flexible delivery mechanisms that seamlessly integrate with modern AI frameworks (these tools have vastly improved over the past few years)
- Consumption-based pricing that aligns with developers' actual usage patterns (see CarbonArc interview below)
- Streamlined tools for AI applications to access and integrate external data sources
- Enhanced transparency and control over how data gets used
- API-first architecture designed specifically for AI applications (this would be great!)

#2 – Thomas Redman and Donna Burbank published How to Make Everyone Great at Data. January 2025.

My Take: The article stresses the importance of everyone in your organization, down to the lowest level employees, understanding the importance of data quality, ultimately creating a culture that values pride of worksmanship.

#3 – Mary Pratt of CIO Magazine published 10 AI strategy questions every CIO must answer. January 2025.

My Take: I love a good Top 10 list. #3, & #10 all deal with adding business value. #6 is data quality. One element that has proven important as my company ModuleQ deploys our AI solution is measuring success (aligns with Pratt’s #4: ROI). What does success look like? This has to be measurable & therefore manageable. This goal should not change as you move through the process, which can be tempting since we are all learning on the job to some degree. For example, we might say that “success” is 20% more client/prospect engagements, as measured by CRM activity within the first 6 months of deploying the tool. More engagement = more deals = more revenue. This can be measured and monitored through the usage period. Measurable ROI.

BONUS: Kris Dyer of Deutsche Bank’s LinkedIn Post citing ModuleQ: “The future of risk management seems to be here. Unprompted AI enables firms to move from process-driven to event-driven strategies, optimising capital allocation and enhancing resilience.” January 2025.

What else I am reading:

Randy Bean’s 2025 AI & Data Leadership Executive Benchmark Survey . January 2025.

Dan Entrup’s Tik Tok On The Clock. January 2025.

Moses Sternstein’s Six charts about AI revenue. January 2025.

Josh Schrenker’s DOGE-y style: The birth of the dearth of data. January 2025.

Dylan Anderson’s What does data even mean? January 2025.

Thani Shamsi’s DaaS is stickier than any other business service - that’s why it’s valuable. January 2025.

Erik Schluntz and Barry Zhang of Anthropic published Building Effective Agents. December 2024.

Shervin Minaee, Tomas Mikolov, and others published Large Language Models: A Survey. February 2024.

Source: Jake Schuster interviews Kirk McKeown - Beyond Silos: Building a Centralized Data Marketplace. January 2025.

My Take: CarbonArc is attacking how data is traditionally sold and used (this is a very inefficient market today). Kirk does a great job articulating his thoughts on the future of the data industry. The key is to “create lift”.
How to Create Lift?
- Be Relevant
- Be Differentiated (more data and better questions)
- Be Accessible
Of most interest to me was Kirk’s thoughts on AI, comparing the advancement of AI to the publication of Black Scholes in 1973. As a result of Black Scholes, over the next generation on Wall Street, there was a whole bunch of innovation which led to lower costs, more efficient markets, etc.
AI will have a similar impact on the use of data … we will see lower costs, more efficient engagement with data, etc.
Carbon Arc strives to lower the total cost of ownership … increasing density and velocity of data increasing value for everyone in the chain.

If you are not creating lift in one these three elements you don’t have a data business.

Create lift in number of opportunities
Increase the hit rate in opportunities
How much you bet against opportunities

Highlights (24-minute run time):

Minute 00:20 – Kirk’s background & CarbonArc
Minute 02:20 – how is this new model of engaging with data happening?
Minute 04:30 – where was the gap in the data world?
Minute 07:30 – Data as it related to sports world (data assets are siloe’d)
Minute 11:45 – lesson from Wall Street (compare to Black Scholes)
Minute 15:30 – if everyone has the data, where is the advantage?
Minute 19:00 – thoughts on creating alpha in time
Minute 22:00 – the opportunities that will become evident at AI creates disruption

Source: Randy Bean’s 2025 AI & Data Leadership Executive Benchmark Survey . January 2025.

Anthropic’s Clio: A system for privacy-preserving insights into real-world AI use. December 2024.

The most common types of conversations users had with Claude, across all languages. The area of the circle corresponds to the percentage of conversations; the titles are summaries generated by Clio after analyzing 1 million randomly-selected conversations.