Request an Invite to the Inveniam AI alpha Release →

February 29, 2024

No One Will Use AI Until There is Attribution for Trading

No One Will Use AI Until There is Attribution for Trading

So Google, or Aphabet(GOOG), lost $90 billion of value yesterday over their Gemini AI 'headache'. The problem is that the AI was giving bad responses, programmed responses to questions evidencing bias in the programming and the LLM (Large Language Model) that was skewed. This is not the first. Wells Fargo had to pay $3.7 billion because they could not provide attribution to the decisions being made by their AI mortgage solution. The apparent bias in the model/programming was providing outcomes that were prima facia biased, when they had to provide an explanation for the decision, ie attribution of the decision factors per loan, they could not. (

No one will use AI until there is attribution for trading. Compute is used at scale in public markets, think high frequency trading and firms like Susquehanna International Group, LLP (SIG), why isn't this penetrating/driving private markets? In the public markets you have referencable data to trade off of, wether trade data, pricing discovery, or company data at EDGAR, the data can be sourced. This is not the case with private markets.

You would think that large compute at organizations like G42 ( which has created the largest privately owned super computer and has just upped their target to 100 exaflops of compute power are perfectly positioned to process massive amounts of private market data to find the signal in the noise. There are only two problems with this, as highlighted with Gemini:

  1. You must have attribution of source data for QA, and prevention of source data hallucination. DLT and zk proofs using Inveniam is the solution for this, and
  2. Data ownership, where the proprietary data set owner does not want to give up ownership to use compute at scale. Inveniam's Federated Data solution also solves this.

MIT-IBM Watson AI Lab, Amazon Web Services (AWS), and Google, need these tools to commercialize AI at scale. This is the virtuous connection of Blockchain, Data at the Edge, and large scale compute (AI), where combined automation of unstructured data through compute, with an audit trail will forever change private markets. This is how all private markets trade digitally, this is how RWA trade at scale.

Rialto Markets Oasis Pro Mubadala Coinbase J.P. Morgan Further Ventures Tokeny Ownera Polymath Binance Centrifuge Deutsche Börse Macquarie Group Cushman & Wakefield Mastercard SEI Mizuho MUFG BlackRock Altus Group SS&C Technologies Apex Group Ltd Retransform, an Apex Group company Abu Dhabi Investment Authority (ADIA) ADQ


Apple PodcastsApple Podcasts