AI in Practice

?AI is underway.??? Investment capital is being poured into Gen AI (ChatGPT estimates a TAM of $1.3T by 2030) while we try to understand the up and the down side on privacy, fraud, and manipulation.? We are still in the early days about the role of government policy, industrial standards, ethics (see: Coming Wave by Mustafa Suleyman)? AI is profoundly more disruptive than earlier cycles of internet, cloud and crypto.

So we move from the intergalactic AI questions to more practical matter of how this can effect our daily work and company mission. ??As an AI investor and operator I have seen hundreds of young AI firms and talked with dozens of global financial institutions.?? Here are some thoughts on an individual use case---four thoughts on how to get into the AI pool and be part of the change.

Pick a use case with teeth

Find the functions or business questions where you have a good? basis of expertise, relationships and knowledge.?? This will help you understand impact later after you have shared your knowledge to data sets, algorithms and LLM

Remember that AI is iterative and communal.? We are not coding with explicit rules and logic.? AI is data driven, coding is rules base.? When I talk with millennials about engineer interviews at Apple, Microsoft and others there is heavy focus on written and verbal communication skills.

Learn how to ask questions.??

I know I sound like an old man but the search engine culture has spawned people that can transact data but don’t think enough about the question nor what the implications are.? ‘Go to the glass’ leads to unintended traps such as shallow thinking and confirmation bias.?? ?McKinsey, Oliver Wyman still tell interns, “be reflective not merely descriptive'.

….which leads to deeper relationships

Let’s take the use case example of institutional research such as one might find in an investment bank.?? This is a revenue driver for the banks.?? Institutional investors pick bankers and traders based on access to researchers.? ?Many clients want access to the analysts as well as publications

‘Investing is a narrative’ ??explains Barry Hurewitz, former Global COO of UBS Investment Research.? He describes the process of creating an environment for researchers with access to stimuli which, in turn, helps them ask better questions which leads to better thinking.??

This contrasts with the pre-AI world where teams of analysts all using the same public platforms such as Bloomberg or FactSet.?? If everyone is thinking the same thing then nobody is thinking.

?…which leads to better thinking

The better approach captures large language models sitting on public models.?? One company to watch in this space is Enquire AI.??? They tap rich content which is then further enhanced with neural networks with access to 60,000 experts who are available to talk, email, etc ??They put an expert in your pocket.?? The technology ?(called Network? Pulse) is linking experts, LLM and other sources to promote a better insight on correlations, causation, sensitivities.?? This is the path to augmentation and expansion beyond the routine

I would be interested to hear from others about your experiences making AI come alive in your work.