Not only do the computer scientists of 2017 not fully understand the future of computing and AI that when I have presented this: Central Planning’s Computation Problem – a paper from the Mises economist in 2013. I did have to laugh, and I do not evilly mean this, but we do tend to find professors trying to make absolutes out of unknowns. Let’s have a quick review. Feel free to read the whole thing, but I will try my best to only look at the significant bits.
My caveat here is my known acute bias. I am not a socialist, the end ideals of Socialism (or value) does not sit in my product roadmap, and as such I will never be a fan. However, if YOU want Socialism, it would be good to understand if in the future you can use a computer to associate the market.
After a bit of mumbling about how amazing Mises and Hayek is it performs the statement that the basis of the paper is to give the benefit of the doubt… ooooh I would be happy to hear this
Assumption 1: Utility is compared interpersonally.
To distribute resources, the computer needs to decide who to give the products to and HOW they are delivered for that person to use.
This is an assumption that demand will always outstrip supply but lets give them this.... We can see where they are taking it later.
Assumption 2: Utility has a simple, cardinal,
long story short.... the computer must be able to compare a bespoke utility for each or.... assign a number for prioritisation for each resource for each person to ensure that you can prioritize who is first to get the items.
Back to this not having enough resources again.... Other than that I agree. I mean, when we build any business or system we prioritise the needs of the customers against the value. I.e. if a customer wants 10 and we only have time for 1, we prioritise and balance who gets what.
Assumption 3: A computer with perfect information about
utilities and available resources.
Computer must know everything about the customer.
What a great location in the assumptions to sneak in some crap. The customer themselves have no idea what they want, they have no access to all their info and computers have a better understanding using social media, bank details and your interactions with people to be able to understand your needs and yet.... they would not need all that info to ensure they keep distribution going.
Assumption 4: No production.
Simply the paper doesn't take the actual manufacture into equation.
This is a slight cop out even in 2013. We all know in 2017 that most production lines even in 2013 were already computerised. Using simple algo to determine what to manufacture. It is standard BAU in today's society. Ok, so, we shall stick to price determination and allocation of goods based on the customer's needs.
These assumptions are a good foundation. However, I feel we need to focus our attention on the premise that the computer needs God like powers to make decisions.
Now there was no premise in this paper that was evident enough for my low IQ to understand. However, we can presume that the premise is “Can a computer run a market social economy: can the computer ensure prices of all resources are matched to their supply and demand while maintaining regular stocks and less waste”. In other words, if I want a pair of Nike trainers, I am just going to go and request them, and they should arrive in a right time frame (unlike my pairs which are always sold out)
So do we need god like powers? I would need to seek further explanation as to why economists believe this, in today’s tech standards, we already use data driven systems to drive predicted resources required for future sales. So if today we use customer interactions, sales, social media and crowdsourcing to predict a level of stock required…. why would we not be using the same systems integrated into other systems to make the system more dynamic and efficient??? We wouldn’t need to know everything about the customer anymore so than we do now. I shall await someone to clarify this as it feels like economists are looking for the perfect system and ignoring that our current system has a lot of inefficiencies….. yes I agree it’s hardly an out of stock. Generally, we will over stock and waste resources and thus sell to a cheaper market excess stock.
There is a slight push that the idea of socialism would always be to push resources onto customers instead of customers pulling from the stock piles. It is an interesting concept that is being spoken about but is highly unlikely at the end consumer market. I mean…. it would be cool to one day wake up and go, I need some new running shoes, and a parcel has already been delivered (yes I know we have some very similar companies out there and I could integrate my social media, past purchases of running shoes and use strava to monitor wear and tear….. but I still wouldn’t like suddenly getting a blue pair)
Touchstone 1: Preferences are heterogeneous.
Yes, we have already covered that everyone is an individual and that expo places more duress on the computer system. Again this is still based on the assumption that the customer is always correct….. how often have you wanted something only to get home and go ahh fook.
Touchstone 2: Consumer goods are heterogeneous.
YES YES, FFS. We know there needs to be the choice on the market and tbh the customer should be designing their products…. but until that happens, the business shall design whatever products they want. Each business unit would just input a new output to the model which will allow the new SKU to be taken into considerations. My first butchery was a deconstruction process for animals, and we would just add a new SKU based on a percentage of the carcass, and it would take it into considerations and predict what the customer may and may not want. Literally, butchers can not do this, and that is why they make mince out of going off meat.
Touchstone 3: Current limits on processing power
The assumption on this point is that you would use a centralised computer system instead of using a decentralised mesh network to solve equation problems. In 2013 the paper says the world supercomputer is 233 Peta flops and yet Bitcoin miners in total are now 64’000 peta flops…. so yeah we shouldn’t limit to what an economist thinks is a limit on computer power in 2013. Let alone the quantum computing power that has just been proven viable and would perfectly complete these scenarios in seconds.
1000 members in a society with 1000 products. … I am going to ignore this as it is fundamental.
– I am noticing you are just meant to believe this with no proof. Apparently, it’s millions of years to calculate 6 billion people with 80’000 products.
So, Murphy (2006)’s claim (echoing Hayek) that the system would take “millions of
billions” of equations understated the problem by several orders of magnitude!
This isn’t exactly helpful.
this math doesn’t make sense.
this is gay.
ok i need to go away
ok then a conclusion