So maybe you can also elaborate about why invalid assumptions can be very dangerous when it comes to using chat GPT.

Yeah, so there are a lot of assumptions that ChatGPT use in this particular case.

So I think it will be quite harmful without the verification of the data input, right? You make certain assumptions and there's one that is...

quite dangerous is that ChatGPT actually recommended Elaine to set aside a huge amount of cash without knowing her expenses structure is like.

So I think that's quite harmful because you might...

Even though liquidity or income disruption is something that's of top priority of Elaine because of the illness, but if you assume she set aside... a huge amount of cash and not...

I mean, the amount of cash is not optimised for her long-term goals.

So how does a human advisor come into play then in terms of hand-holding our clients through execution? How can we add value in this aspect?