4 Comments

JK papers dated 16/11/2023 & 30/05/2024 have more on this subject: in particular readers’ comments with JK replies on 30/05/2024. I do not understand the fine detail (knowledge of statistics etc), but JK’s distillation explains principles.

Will use of this AI give an advantage, and if so to whom? Potentially funds which have the AI, the individual investor and the company itself.

In theory Yes, and so funds & others will have an advantage over private investors; but - to help PIs - will some company offer this AI to private investors, perhaps on a fee paying basis? That would level the playing field to some extent.

Will the company in a particular Earnings Call now take care (1) to rehearse in advance avoiding pitfalls e.g. ‘Lengthy discussion’ on peripheral matters, so as to game the system; (2) use its own AI to identify likely specific problem areas?

More to be revealed as we obtain more example Earnings Calls where this AI is used. Very useful to have JK reports.

Expand full comment

Yes, I think that many companies will start to take more care to rehearse and also possibly do dress rehersals with AI. I would if I was in that position.

Also note that it doen't seem that this is a partuclarly cutting-edge use of AI, so likely us plebs will be on a more even playing field since we've all got access to Claude and AI etc. for fee or pennies and can train and prompt by ourselves.

Expand full comment

I think this is not a technique that will give analysts or investors a big edge. I know hedge funds already use similar techniques (probably much more sophisticated than this one) to filter information. the key benefit here is as a time saver and to summarise information. The edge will come from how the analyst/investor interprets that information.

Expand full comment

They say in the army that bullshit baffles brains, but now we know that bullshit can't baffle AI

Expand full comment