My own theory - based on nothing other than my gut feeling - is that the 45p cut brought unwanted attention to the people in the market and this was their way of saying 'nowt to do with me! I didn't ask for this'. Even if you end up trousering the tax cut, it essentially costs you nothing to register your displeasure with it. Many of these people know they are not very popular and a short term gain, such as the 45p cut was likely to be, would only have set them up for even bigger pain (50p?) in the future if/when a new govt came in and had them in its sights.
This is a great article but I mildly disagree with the concluding flourish. I think the 'market view' is much closer to an average or aggregation of the input opinions than an entirely separate view (implying that perhaps it is more than or qualitatively distinct from being a sum of its parts). That doesn't cut against the thrust of the article (that the market output is more complex than any individual observer can comprehend). And this is why I think it is disanalogous to the AI example - the problem with the market is a too much information problem (we can't know all the views/motives/information of the individuals inputting), whereas the AI example might be both/or either a too-much info and a 'too sophisticated for us reasoning' problem. Now perhaps I'm wrong and perhaps all the latter type problems dissolve in to actually being former type problems (the question is, is a super AI more intelligent than us because it has faster/greater computational capacity and therefore can 'brute force' making conceptual leaps by just considering more permutations, or is it also somehow better at making those conceptual linkages and leaps?). But I have a rankling feeling that I wouldn't say the market has a view because it is a blind, mechanistic bringing together of individual human actions. Is that what the super AI is also? Is it just a sophisticated Mechanical Turk? Am I?
One thing I think is interesting but under explored is whether our working definition of intelligence (or consciousness) involve normativity and is that something that can be programme or will emerge in very sophisticated AI?
I have been thinking over this comment for the past few days. I think you probably underrate how weird and hard to understand market views are. It’s impossible, for example, to recreate them by surveying participants. Many market participants themselves use semi black box calculation they can’t completely explain; when you combine many of these together, and not evenly weighted or timed, you can get some strange results. Especially when the underlying driving factors are constantly changing in unpredictable and mostly unknown ways.
I think I do appreciate that market process of turning inputs in to outputs is extemely complex but I think my point was more that it is analogous to complicated natural systems like climates, rather than to 'intelligence' per se, or can be said to 'have a view'. And that is because I think intelligence has to be purposeful (i.e. seeking some objective) and having a view involves making a judgement, in relation to promoting this goal. What philosophers would call teleological and normative. I don't think the market in itself has a purpose and more does it make judgements as to what is good or bad. My point was that however complex, it is blindly mechanistic, rather than an end-promoting, evaluating intelligence.
Hi, I love your ‘faith-based’ comment. Christopher Hope on Twitter reported from the CPC today that culture secretary, Michele Donelan, received a huge cheer from the gathering that they are to replace the EUs GDPR with the British Data Protection plan - focused on commonsense while protecting data privacy. This common sense thing has come up before and fits into what you call faith based evidence-free policy. It’s worth pointing out, plenty of comments on twitter said that anyone doing business in the EU would have to comply with GDPR as well as BDP, ie more red tape.
My own theory - based on nothing other than my gut feeling - is that the 45p cut brought unwanted attention to the people in the market and this was their way of saying 'nowt to do with me! I didn't ask for this'. Even if you end up trousering the tax cut, it essentially costs you nothing to register your displeasure with it. Many of these people know they are not very popular and a short term gain, such as the 45p cut was likely to be, would only have set them up for even bigger pain (50p?) in the future if/when a new govt came in and had them in its sights.
Interesting idea.
This is a great article but I mildly disagree with the concluding flourish. I think the 'market view' is much closer to an average or aggregation of the input opinions than an entirely separate view (implying that perhaps it is more than or qualitatively distinct from being a sum of its parts). That doesn't cut against the thrust of the article (that the market output is more complex than any individual observer can comprehend). And this is why I think it is disanalogous to the AI example - the problem with the market is a too much information problem (we can't know all the views/motives/information of the individuals inputting), whereas the AI example might be both/or either a too-much info and a 'too sophisticated for us reasoning' problem. Now perhaps I'm wrong and perhaps all the latter type problems dissolve in to actually being former type problems (the question is, is a super AI more intelligent than us because it has faster/greater computational capacity and therefore can 'brute force' making conceptual leaps by just considering more permutations, or is it also somehow better at making those conceptual linkages and leaps?). But I have a rankling feeling that I wouldn't say the market has a view because it is a blind, mechanistic bringing together of individual human actions. Is that what the super AI is also? Is it just a sophisticated Mechanical Turk? Am I?
One thing I think is interesting but under explored is whether our working definition of intelligence (or consciousness) involve normativity and is that something that can be programme or will emerge in very sophisticated AI?
I have been thinking over this comment for the past few days. I think you probably underrate how weird and hard to understand market views are. It’s impossible, for example, to recreate them by surveying participants. Many market participants themselves use semi black box calculation they can’t completely explain; when you combine many of these together, and not evenly weighted or timed, you can get some strange results. Especially when the underlying driving factors are constantly changing in unpredictable and mostly unknown ways.
I think I do appreciate that market process of turning inputs in to outputs is extemely complex but I think my point was more that it is analogous to complicated natural systems like climates, rather than to 'intelligence' per se, or can be said to 'have a view'. And that is because I think intelligence has to be purposeful (i.e. seeking some objective) and having a view involves making a judgement, in relation to promoting this goal. What philosophers would call teleological and normative. I don't think the market in itself has a purpose and more does it make judgements as to what is good or bad. My point was that however complex, it is blindly mechanistic, rather than an end-promoting, evaluating intelligence.
Interesting thoughts, thanks for commenting.
Hi, I love your ‘faith-based’ comment. Christopher Hope on Twitter reported from the CPC today that culture secretary, Michele Donelan, received a huge cheer from the gathering that they are to replace the EUs GDPR with the British Data Protection plan - focused on commonsense while protecting data privacy. This common sense thing has come up before and fits into what you call faith based evidence-free policy. It’s worth pointing out, plenty of comments on twitter said that anyone doing business in the EU would have to comply with GDPR as well as BDP, ie more red tape.