A few month sago, I had my AI epiphany moment.
I had been working on an important presentation which often involved a very uncomfortable aspect of life for me: ACCOUNTING. And MATH. The presentation involved having to do some very simple data set analysis (a nightmare still if you are not a MATH person), extrapolating data, doing some projections…all things which I had to do myself because this was my baby.
After working on it for what felt like forever (probably 1 month), I circulated my presentation to a few close friends for comments. A few came back with very simple comments (we live in a busy world) but chiefly one of them said “it’s wordy”
Of course this sent me into a fit of tizzy. I write well – wordy….surely this is normal?
Another friend suggested – why don’t you give it to ChatGPT and ask it to do the deck for you?
Naturally this outraged me even more. Why would I give it to AI to do it? I am not a lazy person. I am capable of doing the deck…I just need some feedback.
After some persuaded, I decided there was no harm in doing it. But I decided instead of using ChatGPT, which I had seriously prejudices, reservations and general dislike for (simply because it was so commonly cited by a very IR/Friend Structure person I knew), I would use Claude by Anthropic.
Sidenote: my severe dislike of ChatGPT relates to the fact that I have a huge distrust for Open AI, and I have checked the BaZI of the founder of Open AI and did not like what I saw. (but there’s more to the whole ChatGPT thing…)
Rage Against the Machine
I have been generally against AI for some time. The shrill ‘sky is falling’ panic that people seemed to buy into because jobs will be lost seemed to not take into consideration that perhaps the problem is not insufficient jobs as just too many inefficient people employed? Or perhaps just a poor match between skill and need?
And what of the ‘our lives will change’ big well? Notice how you cannot talk to your microwave yet – still pushing buttons – so much for AI. Last I checked the washing machine couldn’t even sort colours from whites. The profound change being heralded apparently relates to being able to turn your face into a cartoon character.
AI was also having a profound impact on the world of content creation – a world I know well. I personally felt most of that was rather alarmist – no one will make movies now! Why do we need actors?! Why do we need AI to make videos? Do we not have enough of them already in that giant video junkyard called YouTube? And is there a reason why we particularly dislike having humans perform aside from the cost of their salary?
Finally, it felt like so much hype, and so many people using it but for INCREDIBLY STUPID THINGS. A friend using it to make clever captions for his IG (I mean…really?) and people saying – look, I threw it into AI and it did my work for me. Kids cheating with their homework.
My biggest objection to AI was basically that I saw it as the bastion of the LAZY and UNINSPIRED.
As someone who regards work as SACRED (I am DW after all), it was enough to make me think AI was going to stupid-fy us all and I was going to be the last to sign up for this kind of lobotomy. I like my brain thank you.And also, if it’s doing all our work, what are we doing with our time? I doubt it’s solving the population crisis, if we have one that is.
Sidebar: Period 8 energies?
I mentioned to a fellow metaphysics friend a while back that I was convinced that AI was part of the dying residual energies of Period 8. Period 8 which saw so many things unleashed upon us in the realm of social media by YOUNG MEN (which Gua 8 is associated with). Soc-med we can all conclude has been generally good for consumerism but has proven to be rather bad for mental health I would seem (except for those of us with already formed frontal lobes when social media arrived).
It felt like that hype that was associated with Crypto…Bitcoin…Scams…Financial Fraud…and Birkins – all of the things which dominated headlines in the last 5 years of Period 8 and well, the start of Period 8 if one remembers that infamous 2008 financial crisis (kicked off by what else – subprime mortgages, the classic ‘Black Tortoise’ (if you’re into Qimen) or Indirect Resource (if you’re into BaZi)
Crypto was Period 8’s middle child, and here was AI, the baby of the family and screaming louder for attention than the rest. Exhausting.
I was convinced all this would eventually reach a proper sane level of rationalisation once the Period 8 energies finally left us although period changeovers can often be tediously slow – like we say in the Metaphysics world: it ain’t all instant noodles. In between the extolling of the virtues of being able to rip off Studio Ghibili’s work for your stupid profile picture and people talking about how AI did their stupid boss’s document request, it was all starting to sound very very IR.
Can AI (so IR) actually be DW?
But back to my decision to use Claude by Anthropic to analyse my presentation.
I chose Claude instead of ChatGPT because Anthropic has a greater commitment to privacy and ethical considerations. ChatGPT just felt too much like the choice of the class clown – not an association I wanted to have. Claude matched my aspirations – cerebral, and not prone to endless emoticon use. It also in my mind, had a nicer interface. Classy. Tasteful. Harvard instead of a party school.
These features appealed to my personality because I saw myself as that – a serious person who was doing something serious (not frivolously designing captions for my IG nightly 5k steps)
At first, I argued a lot with Claude. But gradually it felt like a conversation. And before I knew it, I was opening my wallet to a subscription because it was annoying to have my messages gas out.
And I paid because I was seeing how Claude was actually helping me refine my presentation by essentially playing devil’s advocate but also, drawing on knowledge that it had about investment banking, finances, accounting etc (some which I also had but needed help zero-ing on) then forcing me to keep on improving my slides by challenging me to think harder. It also was able to do some incredible things like analyse numbers, throw out some conclusions, and then debate with me about those conclusions (perfect for the HO skeptic who wants to get it right). It ran data set analysis for me and then throw out conclusions. It was like a smart employee I needed to do my low level version of my presentation which I then red-lined.
As I went through the process for 5 days in a row (and entirely loving Claude!) – I had my AI EPIPHANY.
I went from AI contrarian (it’s going to make people lazy and stupid) to AI advocate (it works if you know HOW to make it work FOR YOU). DW => DW thinking but with a dose HO. I also realised that the secret sauce for AI was all down to perspective: the smart perspective of AI vs the Rest of the World’s dog chasing its tail perspective.
But this is a post from someone talking about BaZi mostly so let’s BaZify the discussion with the BaZi perspectives.
Smart AI vs Dumb AI
Let’s use a comparison that just dawned on me – kids using it to do homework.
Kids cheating on homework has been something that has been happening since the dawn of time. In the past, parents did it or someone had an older sibling who helped etc – the reality is crutches have been around forever. We just never knew about it in the past. That’s because it was not a crowing point to say your pet dog did your homework and got all the algebra right. BUT, we now live in an age where FLEX is the new catch phrase.
Flexing of course has also always been around. It’s just the original homework Flex was getting it done and keeping quiet about how but scoring the Flex on the 100 percent score (and secretly knowing you Huckleberry Finn-ed it). The new Flex is ChatGPT did it and I’m cooler than the rest of you because my dad let me use ChatGPT!
Of course, kids don’t know better otherwise they would realise announcing your father is a lazy parent is hardly a flex.
Who is using it? Probably the kids with strong IW/IR in their charts (‘I know this so it’s a waste of time’ or ‘I don’t care about this subject so whatever’) or the parents with strong IW/IR in their charts (‘I don’t want to waste my time’ and ‘school is not that important’ or ‘cheaper than a tuition teacher’ or ‘I already pay for expensive private school now I have to do homework with them!’)
Now, when strong IR people use another IR-driven tool, the result is a lot more IR. Which means using more or your negative element (since you already have strong IR). Strong DMs using more IR is also kind of pointless – you are simply dipping into what you already have and it’s not doing anything for you.
This is what I call DUMB AI. There’s nothing wrong with what the machine is doing, just the USER is wasting time on it. (sound familiar?)
(Also, strong F/RW or strong IR are the ones busy proclaiming how AI is going to change the world – ask to look at their prompts and questions to AI and you will know what is ‘REALLY’ going on).
So what is Smart AI?
In my presentation example, I decided to go on to build a custom calculator to do some of the maths involved in the presentation for me and clients who I wanted to interact with the data points. But Claude kept getting the Google Sheets formulas wrong. After 3 days of slogging and finishing the calculator (still a great outcome given it was ONLY 3 days and alone I would have probably thrown in the towel after day 1), I asked Claude why it got the math wrong. And Claude answered;
” Sequential precision: Math requires every step to be exactly right – one small error early on compounds through the whole calculation
Context switching: I was juggling multiple formulas, cell references, and logical flows simultaneously
Domain complexity: Your waterfall calculations have intricate business logic that’s not just “pure math” – it’s financial modeling with conditional logic”
Bingo.
When he put it like that, it made me realise what we were doing was actually quite hard but also, on my own, I would have had the depth of domain knowledge to do it by myself. BUT, with some understanding and knowledge, I could drive the AI to make it HELP ME do what I wanted.
Smart AI doesn’t assume the machine is always right but the machine is helping find ‘the right answer’. Smart AI is treating AI like your Friend or Rob Wealth Star which you can spar with if your job lacks co-workers, but it’s also your best HO employee combined with your best Resource employee (the devils advocate meets the clever geek who knows everything who you can quiz endlessly and who won’t complain about you being a time suck on them).
Smart AI is using the machine to do the final 10% of refinements (which invariably are the most painful), and the heavy lifting that represents your weaknesses. It is not building your deck for you – it is helping you make it go from B- to A+. And you can only do that if you are HO/EG driven (skeptic + perfectionist) and you are DO conscious (“this has to be correct, not just ‘hand in”) and you like the work (DW!)
Finding the Right AI for your Chart/Structure
Broadly speaking based on a brief test of 4 platforms (Gemini/ChatGPT/Claude/Grok via Twitter), it’s possible to briefly approximate what kind of AI platform suits which kinds of Chart Structures although I want to caution this is a work in progress by way of experimental observation.
Strong HO types/7K types without the moderation of Resource will be drawn to Grok as they appreciate its non-woke tone and shoot from the hip style. Grok also likes to ask for forgiveness, not for permission (like its founder, Mr Musk). So the notion of how Grok works is that it can say x, then change its mind and just apologise.
Strong DO types will likely appreciate Gemini more because Gemini is more centred on working with documents like Google etc. Gemini feels like the original WORKHORSE – auto drafting those boring emails, reading your excel sheets, finding you stuff in your Google Drive…the original secretary I suppose. I’ve found it’s engagement for other things rather mundane despite the information sources likely being the same. Perhaps because Google Search has actually been crappy for years we just never realised it. But apparently it does meeting minutes really well (waiting for a test answer on this) – again, how secretary!
Strong HO with DW types or Strong HO with DR, or HO/EG will prefer Claude in my view. Claude is very cerebral sounding, has a classy interface, can be pushy without being in your face, and not too grovelling. Claude also gets anxious about copyright which can be annoying if you know you’re only just doing a ‘WIP’ item not something to be circulated to the world. Claude is not a cheerleader – just that smart, sensible colleague who helps you out but doesn’t DO YOUR WORK.
And at last, the popular favourite: ChatGPT. Personally I use ChatGPT for nutritional tracking and I have found it to be much better and more effective than Claude (which I pay for!). ChatGPT tends to be more energetic, ra-ra and attaboy compared to Gemini or Claude. Very good when you ate pizza for lunch and now need to know how to balance the macros for the rest of the day. Also, helpful and very very supportive at the same time. All the RW and IR of a sympathetic personal trainer without the DO food judgment! (and thinks faster than your PT any day)
So, if you have a lot of RW/IR in your chart, or intense IW, my view is you should AVOID ChatGPT as all it is doing is playing to your chart’s dominant elements. And that is not how you should use AI – use it to challenge your assertions, not reinforce your chart’s dominant aspects (and technically what you don’t need more of anyway!)
This hypothesis may change as more platforms come on line and more start-ups try to join the so called AI revolution – I reserve judgement on ‘revolution’ point. But in terms of its capacity to help with getting work done (a very DW efficiency point), it remains subject to the ability of the user to actually use it effectively with correct prompting which is actually a HO/EG attribute (HO is best at asking questions, EG for asking deep questions).
Bottom line: the Fundamentals of the Universe NEVER CHANGE.
An integral aspect of surviving today’s world is adaptability but not just pure adapting, but a balance of adapting (not reacting) and thinking (not just doing). I had been trying AI platforms for a short while before my big epiphany weekend but never using it actively as I have in the last week. (and no, this post was not written with Claude’s help).
As someone who is good with words, I did wonder if the machine could exceed me. But in my interactions with Claude I realised – the machine does not exceed me. It is for me, to exceed it. Not just cave to the 70% percent option it throws at me. (also it’s lousy at Metaphysics – that’s for the next email!)
The absolute basis of all Metaphysics is three things: the FIVE ELEMENTS, the COSMIC TRINITY (heaven/earth/man) and the most fundamental of it all, YIN AND YANG.
Nothing in this world is YIN OR YANG. It must be comprised of both, just in different ratios, and forever in a necessary battle for the perfect balance.
It’s the same with AI.
As soon as everyone starts using it for homework or grunt work, something will happen to re-set the balance. And let’s not forget, this now Period 9. All those Black Tortoisy-IR types that thrived in Period 8? The sun has set. And what may be coming out right now are just the final vampires.
So those lazy ass employees and students who are using it to do ‘homework’ or fulfil their KPIs – the chickens will come home to roost sooner than they think. Because if there is anything the universe abhors more than a vacuum, it’s an imperfect balance.
📩 Subscribe to the BaZiBuzz Newsletter
🛒 Buy My BaZi Courses