Discussion Forums » Big Issues of our Time

Will AI etc. mean less hours and less tax revenue?

    • 23 posts
    June 29, 2021 9:26 AM BST

    Whilst I'm all for working less hours which AI might lead to - I'm in a bit of a quandary. Surely if we work less hours (e.g. Double-jobbing like Canada) the government will get less money? I can't see how the sums work out. Whilst increased 'leisure time' might be welcomed how will some people afford the lifestyle they want if they work say 20% less hours or a 4 day week instead of a 5 day week? I'm thinking of the consequences for loans, mortgages and buying power (and hence VAT). We might be the agents of our own demise - be careful what you wish for, you might just get it. Where's Len Murray when I need him? Plus is society really ready for the lessened 'sense of well-being and purpose' that the lack of jobs for life might herald? Plus why do the govt. (through OA pensions) give out leisure money near the end of life not beforehand when it would be more needed? This is in addition to what was in the Manifesto from February.

    This post was edited by Chris Barnes at June 29, 2021 9:31 AM BST
    • 102 posts
    July 6, 2021 10:10 AM BST

    Chris you raise some important points.

    The key piece in the equation is that AI leads to increased productivity and not just small increases, potentially order of magnitude ones. Productivity increase is the one thing that allows companies to make more profit and hence HMRC and also enables companies to increase wages. So with substantial increases in productivity everyone can be  a winner. In particular it opens up the possibility of reducing hours but paying the same wage.

    Note that at one level this is part of achieving the 'good of all'. At another level it is going to be necessary if we reach a job tipping point due to AI where we start seeing a net loss of jobs. 

    Paying for this is all part of our Productivity Contribution mechanism covered in our Manifesto of Ideas and in our series of AI documents

    Artificial Intelligence - Managing the Future (This introductory document to the topic)
    Artificial Intelligence - Planning for an AI Society (Detailed Scenario Planning of alternative potential scenarios)
    Artificial Intelligence Society - Proposals (Additional proposals drawing on the Scenario Planning document)

    Also for a short overview of the top items take a look at the document just posted at AI and the Impact on Society which is something we have raised with the All Party Parliamentary Group for AI and also the AI Council because we believe this and other 'impact on society' things are not being addressed. 



    • 23 posts
    July 6, 2021 10:56 AM BST

    Maybe it's the sceptic in me (and it wasn't AI related) the BBC today reports on the success of a 4-day week in Iceland - what was in the small-print is that it was mostly LG positions concerned and they paid a 5 -day per week salary - privately held companies may take a different stance. If the 4-day week helps productivity then as Tony sort of mentions WRT AI this may help business see the 'gains' even if it's not purely AI related. All this said, moves towards a shorter working week may act as a catalyst for AI being introduced to wider spheres than are currently envisioned.

    Once again the coding sceptic in me says, if AI becomes more prevalent how will 'version control' work and how will we prevent any 'malpractice'? "I'm sorry Tony, I'm afraid I can't do that". Funny in 1968, still funny in 2021? Mmm.

    • 102 posts
    July 6, 2021 2:16 PM BST

    I think productivity gains have to lead any reduction in working hours not vice-versa.

    As far as version control it's a good point because the risk associated with AI has to be managed throughout the deployment and maintenance life cycle. Some of McKinsey's reports are very good on this.

    As far as malpractice there is a big debate on AI ethics and regulation including human overseers, explainability of AI decisions etc. We've recently done some thinking on this ourselves at AI - Ethics and Regulation