Investment

New York and Toronto-based family offices back next generation AI platform, so-called “casual AI”

Family offices have agreed to back a next-generation AI system developed by causaLens which is capable of making decisions based on data analysis.

One family office investing in its new $45 million funding round is Dorilton Capital, linked in the media to insurance tycoon Calvin Lo. The relationship has not been confirmed. Dorilton is viewed as a savvy niche investor under CEO Darren Fultz, ex Rothschild Inc. Its ventures include Traditions Health, a hospice operator; the Formula 1 Williams racing team; and the Midwest cooling tower business.

Another causaLens investor is the venture arm of Generation Capital, a Canadian family office owned by Graham Beattie. The three other investors in causaLens are IQ Capital of Cambridge, Molten Ventures and GP Bullhound, who have each developed a reputation for backing deep tech.

CausaLens says it has the only AI application which scans data to develop causal links which can inform human decisions.  

The approach is called causal AI. It has clear applications in healthcare, as illustrated by ClosedLoop.ai of Austin, Texas, which wants to define the causes of disease. It is backed by several investors including Jim Breyer of Breyer Capital, a renowned backer of US and Chinese entrepreneurs. 

Until now, AI has relied on machine learning systems that teach themselves to develop conclusions based on an analysis of past events which may no longer be relevant.

Realtor Zillow, for example, came unstuck with machine learning in November 2021 when it was using the technique to match house buyers with sellers, only to suffer a systems failure when its algorithms misjudged a price slowdown. Zillow saw 45% wiped off its market value in a week.  

CausaLens’ co-founders are chief technology officer Maksim Sipos and Darko Martovski, a former hedge fund researcher. He says machine learning often makes mistakes through ‘overfitting’  when false conclusions are drawn from past experience, often from sets of data that are too narrow.

In contrast: “Causal AI achieves high predictive accuracy by abstracting away from features that are only spuriously correlated with the target variable, and instead zeroes in on a small number of truly causal drivers.”

Aviva Investors portfolio managers use causaLens, saying the relationship has led to the development of: “new causal relationships in economic, financial and alternative data, with sophisticated, adaptive and alternative models that don’t suffer from overfitting.” 

CausaLens has produced short-term data which indicates causal AI can produce superior risk-adjusted returns.

It is advising systematic hedge funds, hitherto reliant on machine learning ‘black box’ approaches. It has started an insurance business, to inform decisions in underwriting and capital markets.

Nuveen Real Estate has used its system to improve its decision making in areas like inflation-modelling and post-Covid 19 modelling.  CausaLens can model patient flow in healthcare. It seeks to reduce inefficiencies in supply chains and energy supply. Clearly, it has no shortage of ambition.

Breakthroughs, including causal AI, are taking place across supercomputing.  Meta Platform is close to creating AI with a range of capabilities, such as translating hundreds of languages in real-time. Alphabet’s DeepMind is seeking cures for the world’s illnesses. IBM’s Watson wants to solve problems for businesses.

But Davids often give the Goliaths of the tech world a run for their money, and causaLens is unlikely to be an exception.

Subscribe

You will need a Premium Plus Subscription to access this database.

Exclusive news, analysis and research on global family enterprise and private investment offices.

Access to the most comprehensive fully interactive database on global family offices, principal investment offices, and family enterprises.

Check Deal Data, Senior Staff, and New Analysis on more than 1000 family/principal investment and holding groups

Already have an account? Login

Subscribe

You need at least a Premium Subscription to read this article.

The most comprehensive information service on the global family enterprise world, featuring exclusive news, analysis, research and data on global family enterprises, family offices, and private investment offices.

Premium

£299

per year

  • Exclusive reports, analysis and commentary
  • Exclusive access to family/private investment office deal information
  • Exclusive interviews with principals and senior management of family/investment offices
SUBSCRIBE NOW

Premium+

£399

per year

  • Access to All of Premium
  • Access to all of FamilyCapital Analytics, our interactive database with more than 1000 detailed profiles of family investment groups

More Info

SUBSCRIBE NOW

Already have an account? Login

You've reached the end.

Continue reading free articles by registering as a Member.
Or choose a Premium Plan.

Membership

Free

  • Exclusive reports, analysis and commentary
REGISTER NOW

Premium

£299

per year

  • Exclusive reports, analysis and commentary
  • Exclusive access to family/private investment office deal information
  • Exclusive interviews with principals and senior management of family/investment offices
SUBSCRIBE NOW

Premium+

£399

per year

  • Access to All of Premium
  • Access to all of FamilyCapital Analytics, our interactive database with more than 500 detailed profiles of family investment groups

More Info

SUBSCRIBE NOW

Already have an account? Login

Leave a Reply