Some really good points there @Vandelay

I likewise think that it's going to build any moat based on the capability of the software alone. You need the added elements of network effects, proprietary data/IP, scale etc if you're going to fend off AI successfully.

@Vandelay good issues to raies. my 2c on this is that it hasnt changed that much. I recall a paper from Stern Uni that ranked the longevity of moats for each industry and tech was the lowest. (btw utilities were the highest ) both are unsurprising to me. I think AI just extends that into the future. what always intrigued me was the markets willingness to price the growth element much higher than the element of disruption or uncertainty of longevity in tech stocks. i feel that won't change. the winners from the mobile revolution will probably pass the baton onto, yet to be identied winners int the AI space. will it only be MSFt and GOOGL? probably others will arise. for those with a lack of defensive moats like IP, scale etc it is just as uncertian as it always has been to me. My disposition is for quality profitable growth companies and that won't change, but have to keep a weather eye on the risk of obsolescence, as always as it evolves. that my view

@Vandelay Interesting thoughts and questions, and ones I have spent quite some time thinking about this year - and will no doubt continue to think about in future.

My thinking on this is that software and "internet" companies that codify and internalise business processes have a relatively strong moat. This is because they require well-tuned human and system process and data interfaces with the "users", where there uses can span a value chain. The human interfaces require training and behavioural adaptation in order to be effective, and it is the behavioural adaptations that are hugely important. Equally, there must be a level of trust by staekholders in the process and the tech. that a given set of inputs will lead to predictable outputs, with clarity on what is "prescribed" or certain.

Through their integration into mission critical processes, most software and many "internet" entities gain access to and control of a huge dataset beyond pure data and into value-adding information into the wider business evironment that can be leveraged. So the question for me is which firms position themselves to exploit this opportunity and deepen their moat around it?

So I therefore expect that software and "internet" firms will not necessarily become poor investments. On the contrary, those that make appropriate investments in ML and AI exploitation will deepen their moats and become even better investments.

So, I value highly software and internet companies that maintain a high level of spending on R&D as a % of revenue, and where they provide practical examples of how some of that spend is going into exploiting the huge datasets they hold.

This is why I maintain a high conviction on firms like $ALU, $TNE and $WTC, and its why I am giving firms like $VHT and $M7T a go.

However, from a portfolio management perspective, I also want to hold firms that deal with physical stuff. Hence I hold firms like $RMD, $FPH, $CSL as well as $LOV, $NCK and $SGI.

Your questions are great questions because that help us evaluate systemic risks in our portoflios.