I’ve been prompted to write something on this subject by a couple of videos that popped up as recommended on my YouTube feed this week – although it is also in many ways following on from my previous post when it comes to being well informed and the Dunning-Kruger effect. The idea, of course, is fundamental to the purpose of this blog – i.e. to continue to learn and to constantly remind yourself that there is much yet to be known.
The first video reminded me about Plato’s allegory of the cave.
The focus for me would be on the fact that when the ‘enlightened’ one returned to tell the other captives of what he had learned they did not have the capacity to believe him. This is exemplified frequently by people who “used to be” MAGA trying to convince those who still are that there is a big world out there.
The second video was the ‘banned’ TED-talk by Rupert Sheldrake where he pointed out the fact that scientific progress is seemingly often hindered by what he termed dogmas that scientists refuse to entertain may be wrong.
Specifically, people always assume that what they already know is “truth” and are very unwilling to change that ‘knowledge’ which results in confirmation bias kicking in and an ignorance of anything which contradicts your worldview.
This all combined in my thoughts with an interview with a young entrepreneur on Bloomberg TV in which an ‘interesting’ picture was painted of his knowledge. I get it that there are some people in every generation who are well ahead of their peers – you see plenty of examples each year around A level time when a (very) few people score exceptional results – and it is newsworthy because it is not a common phenomenon. I could relate a whole host of such examples through the years, but the point I am making is simply that these people are rare.
They excel (albeit in a limited environment) and this is great.
The “Internet Age” for want of a better word has, in contrast, thrown up a huge number of “influencers” – or whatever they may wish to be called – and an even greater number of young people who believe that they are extremely well informed. The particular young man interviewed on Bloomberg is not someone I have any knowledge, so this is neither criticism nor praise of him – simply comment. He came across as supremely confident, seemed to “know his stuff” although in the short time I watched the video I didn’t actually figure out exactly what “his stuff” was. In short, he seemed typical of lots of people (mainly online) who appear in our YouTube feed or other such source and “sell” you their view of whatever part of the world they know a bit about.
The issue is that – usually – that ‘knowledge’ tends to be pretty superficial and rarely looks beyond the ‘immediate’ in terms of how their ‘advice’ helps. This particular guy did seem to have more than the usual depth to his knowledge, but in terms of “those like him” I would say he is most certainly in a minority. The vast majority suffer at least in part with the Dunning-Kruger effect.
Now looking more specifically at the current situation with the “search for inefficiencies” within the US Government agencies we meet precisely the same sort of thing. I don’t doubt that hidden in those agencies are valid examples of genuine waste, whether through incompetence, honest mistakes, lack of resource and, probably, some deliberate fraud. However, I would posit that the idea of sending in team of people who have no background in the ‘context’ (or perhaps contexts, since it is definitely possible that there are many competing and conflicting factors) is an immensely stupid idea for what are they actually looking for. If it is “waste of money” – what does that mean? If it is “paying too much” – how do they arrive at what is reasonable?
The questions go on and on. A further – and VERY important consideration is how do you actually go about measuring inefficiency or ineffectiveness and at what point does something “cross the line” into those categories. You can imagine that it is likely that they will find the low hanging fruit rather than the significantly ineffective. Of course, much of government spending is also multi-layer with contractors and sub-contractors and sub-sub-contractors and so on – how are the inefficiencies in that chain going to be identified?
All that is before you get into the issue that not everything can actually be measured – a huge proportion of this is necessarily objective – whether someone qualifies or does not qualify for a benefit for example will have a mix of objective, measurable parameters and a lot of subjective, “gut-feel” parameters. Yes, you can put numbers on them but the assignment of those values will itself be a subjective process.
Then there is the cost of not spending the money – that sounds like a perverse statement but is often very true. Thought experiment – there is an illness that is expected to affect 10% of the population and will result in a significant medical cost for each person affected (although it is not known the distribution of “seriousness” of the illness. There is also a drug which – if administered before infection – will protect almost everyone. In this case it is sort of clear that the cost of not spending the money to provide the drug is the relatively unknown cost of treatment. How on earth do you measure the efficiency here? (Of course, this simplifies the real problem since there will be other benefits and costs which are relevant.
It seems to me that throwing a team of (and here I am guessing at their skills) data miners at this problem is a monumentally stupid idea – assuming that what you are REALLY trying to do is root out the inefficiencies. Remember DIKW? Data is, on its own, not terribly useful. Simply saying “they spend $1m on xxx each year” is not a helpful statement – worse it is a misleading statement depending on what “xxx” actually is.
Here is an idea 😉 – it was reported that the cost of President Trump going to the Superbowl was around $20m – which made me think – how much is spent on ‘supporting’ the president during his four years in office? How about changing things so that the president never needs to leave the White House for four years – how much would that save the country on transport, security, ‘wasted time’ both for the president and for those who ‘need’ to follow him around – there is so much more. Strikes me that – ignoring all the subjective and contextual aspects this would be a huge saving and wipe away a lot of extravagance.
OK – so that extreme may not be possible but I am sure there are many ways in which much of that inefficiency could be wiped away.
So – coming to the title theme of this post – the team that is enthusiastically gutting US government spending seems to be composed of “youngsters who know” – although what they know is still a bit of a mystery. Whatever it is, it almost certainly does not include the mechanics of – for example – military procurement. It won’t include the ‘soft’ effects of USAID’s spending around the world – they may never have entertained the fact that this was something that mattered. It won’t include virtually everything that isn’t how to find data inside a computer’s memory.
There have been reports of the use of what are effectively keyword searches – this is not by any means a foolproof technique. AI has lots of plus points, but a significant defect is that it has no way of correctly factoring in all of the variables that make up a decision. Think for a minute about sitting down at a lovely restaurant with a menu packed with your favourite dishes. Which one do you choose? Why did you choose that one – and go beyond, because I fancied it at this time! There are likely a number of (seemingly) unrelated things in your brain that led to the choice – and not just in your brain. What about if your companion chose a dish you were thinking of having, but then you chose a different one so there was variety – now it is also all those things going on in your companion’s brain. What if there is a group of eight or ten or more – lots of brains to consider now.
So, the decision to send $1m of aid to some middle eastern country may be an inefficient use of those funds. However, think about the huge number of factors that may well have been considered in arriving at the decision to do that piece of funding. Of course, one of the options is that the decision maker took a bribe – that would make it not only inefficient but also illegal. However, I have to believe that the vast majority of such decisions are made as a result of the best knowledge available at the time.
Time – brings up another thing here – looking at past decisions it can be easy to point to those which did not turn out the way that was expected – so, with today’s knowledge it was inefficient or ineffective. Decisions are made at a point in time and I along with others have stated before that you NEVER make a wrong decision because whatever decision you arrive at is based on all the knowledge you have AT THAT TIME. You may well discover a few seconds later that some of your knowledge was incorrect or incomplete – that doesn’t change the fact that at the time the decision was the right one.
This is an important point for two main reasons – the investigators later cannot apply “better” information to your decision and just because someone doesn’t like your decision they cannot say it is a wrong one, because their “total knowledge” will be different to yours resulting in a different interpretation of what the decision should have been.#
In turn, we come back to the (dis-)information idea from the previous post – if you are basing your decisions on incorrect data then inevitably your decisions will be skewed – they will still be the ‘right’ ones for you to make. Perversely therefore the DOGE team gutting the US government are making the right decisions – which only goes to show that it is vitally important to get the right decision makers – which, perhaps is another long essay!!
The fact that the DOGE team is young is a huge issue, however we who are older also need to realise that we are always too young to know everything. This echoes Sheldrake’s points regarding science where they believe that because they know it is no longer necessary to keep looking – never fall into that trap – whilst the unknown unknowns are without a doubt a huge contributor to any risk – a much greater danger are the knowns that are simply incorrect.