Pick a number any number
Pick a number = any number
I was very pleased to read an article by The Rt.Honourable Steve Baker which talked about how bad outcomes of ideological policies are.
In essence, what appears to happen is that some numpty thinks a good wheeze is to (for example) tax sugar in soft drinks. Now you would think that some research would go into how much is used now, how much taxing it would reduce it, what might be raised etc etc. In reality, none of that happens. The rule of thumb appears to be how much is used now (and personally I’d be very surprised if that was an accurate figure) take the cost and let’s say the idea is to reduce that figure by 20%. No input to prove that or even work out if it’s possible. So those working this out (HA!) suggest a tax of 20% should do that. Which, by a process of cod mathematics equates to raising say £2 billion in tax.
What happens, of course, is different. The moment you interfere in a market you change the dynamics and human beings are nothing if not adaptable and market-driven. The demonstration of this is particularly striking. The announcement is made that the wonderful new policy will reduce sugar in drinks by 20% and add £2billion to the treasury. Win,win. Unfortunately, our present number crunchers have no idea at all about market dynamics.
People get in the way. Guess what? They take avoiding action. The three policies I particularly like are highlighted in the article. 50% tax rate will raise £2.7 billion a year. Except it didn’t. It was barely £1billion – and falling. Energy Profits Levy – over £41 billion. Nice chunky number. Ended up around a not-so-chunky £17 billion and stopped energy development in its tracks. So also going down. If ever there was a case of shooting yourself in both feet this was it. Non development = no workers = no tax income = unemployment benefits paid out. And finally a real doozie. Cancellation of non-doms. Filthy rich people- tax them! Raise £34 billion! Guess what -they up and left to such an extent that in all probability nothing will be raised at all. And again employment suffers, tax take suffers, and other countries benefit. We are terribly good at that part.
Now here’s a thing.
One of the absolutely certain ways to make money is to spot when something is mispriced. Priced too high and you short it, priced too low and you buy it.
So I was very interested to read an article relating to crypto and AI which was promulgating the idea that crypto was underpriced in relation to AI.
I’m not sure how you could be certain about this but what I do know is AI as it stands isn’t making anyone rich. In fact, if anything it is making some extremely wealthy companies poorer. Just so you know, crypto is doing exactly the same. I know of two listed companies that – despite the rise from say $30,000 to $126,000 in Bitcoin over the last 3 years - have managed BOTH to lose money. I don’t pretend to know how they have done this, but do it they have. Even Micahel Saylor has managed to “lose” billions.
There are quite a few Bitcoin mining companies that have figured out that the game has changed and now is the time to pivot (a lovely old word that denotes giving up) away from mining. As I’ve said before mining farms already have most things in place to be data centres. They have the cooling, the power and racks to put the hardware into. You have to get rid of the mining rigs of course as in actual fact they are only any use for mining and nothing else. But throwing them out and bringing in shiny new computers to put in the racks is a very easy job. And they will probably earn more money than the old Bitcoin mining rigs did.
What is interesting about AI is not so much the hype but the extent and breadth of it. Has anyone reading this managed to get the chatbot of some mammoth online company to actually answer the specific question being asked? I have to say I haven’t. As recently as today I received an email suggesting I used a service priced at £49.99 + vat – click here. Which I did.
Only what came up was a price of £75.99 + vat.
Now I suppose that some people would say oh well, that’s still not a bad price and they might pay. But – as you may or may not know – I am Scottish and I’m afraid I adhere to the principle that a quoted price is sacrosanct. So I tried for a phone number (no) and had to fall back on chat.
So I typed “Your offer of £49.99 when clicked on shows up as £75.99. Is £49.99 no longer available?” To which the response was. “Here are a list of FAQs” none of which had anything to do with my question.
I won’t bore you with the next 40 minutes or so, but suffice to say I eventually managed to get a human (the answer when I first asked for one was “ I don’t understand the question”) who said they had a glitch and here’s a link to £49.99. Simples.
Anyway, all this is just a longwinded way of saying there is a definite mispricing of AI. I’m pretty sure you won’t have read the terms and conditions of the use of any of the systems, but did you know ChapGPT’s terms and conditions state it is for entertainment only? Unlikely to replace many people short term on that basis, I’d say.
The costs of development and the data centres are so massive, I’m afraid I can’t see how anyone is ever going to make any money out of it – which, let’s be honest, is what companies should be trying to do. I’ll be interested in a year or so to see how accountants work out the value of what these companies have. That said, it is currently incredibly cheap to use, although I would argue it isn’t value for money in many cases. But cheap it is and can be used. The simpler the task the better the outcome. At the same time, I would suggest that Crypto is overpriced. Don’t forget there is absolutely nothing backing any cryptocurrency. Yes I know there is the network, but is that really worth actual cash? I would argue it isn’t. Stablecoins are a different proposition and for sure one of the reasons they have grown so much is that they are effectively a proxy for cash and move at lightning speed. But crypto is merely the result of who will pay what to buy something as opposed to who is trying to sell and get out. Crypto suffers massively from the cult of buy for £1 and sell for £10 overnight when in reality minimal numbers of people actually make money. But the stories of that very thing drag in more people to risk their heard earned cash.
So despite the volatility – which itself argues for a mispricing – I’d tend to be a seller of both AI and crypto. Until I’m not.
