AI forecasting/insight

Long time watcher/investor back in from 2017 days. First time poster.

It's been a long journey to say the least, and as we sit on the precipice, I've noticed the discussion regarding AI. So I'd like to start a thread seeing what people are getting from various AI platforms. I mean we probably don't have access to sophisticated platforms such as Kobold but hey if AI is good enough for them to use...

Since acquiring an Gemini AI equipped smartphone, I've been playing around with the current scenario to see what it can do. I'm no miner or have expertise in the field so I'm potentially missing insight in my approach and the legitimacy of the output. So I'm just offering my non expert opinion.

Whilst I do believe it has potential, the "garbage in garbage out approach" stands strong and yes one can tailor their prompts or the information pool for extraction can be biased.

Ive been trying to feed in as much information in the way of asking for low, mid and high range estimates of NPV combined with updated DFS valuing what's in ground, factoring in associated costs and revenue of the resource over the mines lifespan. Then I'll keep prompting in premiums and discounts considering macro and micro environment factors coupled in with the geopolitical situation tied with Western nations looking at decoupling from Chinese supply lines.

Throw in other potential bidders including the Chinese wanting to retain interest and it gets interesting. I ask for conservative numerical estimates and factoring in all risk and even the "human factor" whilst getting it to cite real world historical examples all the way through and all I can say is I'm impressed.

What I do find interesting it actually acknowledges valid talking points and interesting concepts in considering the variables at play. Personally I value that when prompted, it will show how it arrived its conclusions and cite relevant real world historical comparisons in drawing its conclusions before revaluation of its previous estimates and probability. It will also justify the price ranges and why it would/wouldn't happen.

Using the damages claim as a baseline for valuation. Ive generated dozens of pages with it all sourced on how it got to its conclusion as each added layer of details causes a re-evaluation of its low, mid and high range take over prices.

The assumptions I run with are. AVZ gets its rights to the tenements (75%), the minerals security deal goes through, and its a take over. I do run and prompt for different scenarios such as JV etc and its likelihood and jv and take over are up there.

This what Gemini AI thinks based on inputs.

For a takeover, the AI thinks $3 usd is an absolute bargain. With the lowest offer range at 4 to 6 usd, with a probability of 10 to 15%

Mid bracket is 6 to 9 usd at 60 to 70% probability

High end is 9 usd to 15+ at 20 to 30% probability (Bidding war and confidence in rising mineral prices/strategic importance)

I've spent a couple of hours at a time and gotten similar opinions and results, this round was my most conservative results. (Yes it cam be lower but that is feeding it the bare minimum on information). I do feel engaing in too long of a thread can perhaps trip up the AI (it may forget something mentioned earlier) Im no expert to say thats 100% true just a gut feeling from my experience.

Whilst the AI considers the catalyst mainly being the peace deal going through as a big sovereign derisk and security followed up by AVZ being successful in court for ownership (I would assume a premium would be negotiated to not have it go to court).

Whilst I will patiently wait till its all wrapped up and take it all with a grain of salt, not wasting too much time speculating (although i acknowledge im technically doing that) as nothing is guaranteed. AI is just a tool after all. I will be wishing luck to all holders and AVZ, as this is the chance to benefit stakeholders if done right.

What are other peoples experiences with AI and what have you been using?