Introducing Signal Codes
This is the first tutorial published on the website considertheevidence.org.
Even though it is published here, it is not generated here. Consider the Evidence is a publishing platform - like an online newspaper - but it has very strict standards as to what can be published.
Initially, only tutorials from another site, Signal Codes, will be published here.
Signal Codes is a platform which encourages users, when observing ideas, to thinking carefully and, when expressing ideas, to get to the point.
There are also some rules around ensuring accuracy.
Signal Coders are a tribe of people who use signal codes to understand and communicate their main ideas concisely and precisely.
When Signal Coders approach a topic they do so assuming that other people's ideas deserve a fair hearing. They playfully yet respectfully engage with those ideas, using ordinary narration interspersed with signal codes.
Asa Jomard: AI Dividend
With that in mind, I am going to explore a recent Substack post written by my friend Asa Jomard. She called it, "The AI Dividend: Taxed, or Spent as Time?"
There is no special reason this particular post was selected - Ms Jomard has interesting things to say, and this happened to be her most recent post.
The post focuses on the issue of AI: Since AI will disrupt the job market, how can we ensure everyone still gets a fair income?
Having selected a source of information, my job as a signal coder is to highlight quotes I find interesting.
I found about ten of them, but I won't show you them all.
Why not? There are two reasons:
- Just because I find them interesting, it doesn't mean you will. I will only show you what I regard as highlights.
- They are not my quotes, they are Ms Jomard's. While people give little regard to copyright nowadays, and while the Signal Codes allows you to maintain your own copy of source information for study, it is very strict about how much of that information a signal coder is allowed to publish.
So, let's look at a highlight. Here I have extracted a quote from the start of Ms Jomard's post where she asks where we want to be ten years from now. Projecting into the future seems like a good way to get some useful results: ie. Let's work out where we want to get to, then work backwards from there.
Imagine two versions of the same country, ten years from now, after AI has delivered the productivity gains everyone currently promises it will. In the first version, people still work roughly the hours they work today, but a robot tax and higher capital taxation fund a generous safety net, retraining programs, and expanded public services. In the second version, people work three days a week for the same pay they earn now, because the gains were captured not as tax revenue but as time nobody has to spend working anymore.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
Imagine two versions of the same country, ten years from now, after AI has delivered the productivity gains everyone currently promises it will. In the first version, people still work roughly the hours they work today, but a robot tax and higher capital taxation fund a generous safety net, retraining programs, and expanded public services. In the second version, people work three days a week for the same pay they earn now, because the gains were captured not as tax revenue but as time nobody has to spend working anymore.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
Initially, the author's thinking reminded me of Edward de Bono's C & S tool.
C&SWhat are the immediate, short-term, medium-term and long term consequences of this action?Credited to Edward de Bono
Read about it at debono.com ↗More on SignalCodes →
Imagine two versions of the same country, ten years from now, after AI has delivered the productivity gains everyone currently promises it will. In the first version, people still work roughly the hours they work today, but a robot tax and higher capital taxation fund a generous safety net, retraining programs, and expanded public services. In the second version, people work three days a week for the same pay they earn now, because the gains were captured not as tax revenue but as time nobody has to spend working anymore.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
But then I thought that Asa might not so concerned with anticipating future consequences as comparing two scenarios.
There is no pre-made concept for "comparing two scenarios", so as you can see here, Signal Codes allowed me to create my own personal code 'compare-two-scenarios'.
compare-two-scenariosPresenting two ideas for comparison, not necessarily to choose one over the other, but to think through the factors involved.Read full concept →
Imagine two versions of the same country, ten years from now, after AI has delivered the productivity gains everyone currently promises it will. In the first version, people still work roughly the hours they work today, but a robot tax and higher capital taxation fund a generous safety net, retraining programs, and expanded public services. In the second version, people work three days a week for the same pay they earn now, because the gains were captured not as tax revenue but as time nobody has to spend working anymore.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
The author's first proposal seems simple enough...
if productivity rises because machines are doing more of the work, redirect some of that rise into public hands through taxation, and redistribute it as services, training, or direct support to the people the machines displaced.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
...tax and spend.
tax-and-spendTake money from those who've got it and give it to those who need it.Read full concept →
if productivity rises because machines are doing more of the work, redirect some of that rise into public hands through taxation, and redistribute it as services, training, or direct support to the people the machines displaced.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
She points out that taxing money can be ineffective - distributing it can be slow and not go to those who need it most.
But it depends entirely on the state successfully taxing gains that increasingly belong to owners of capital who can, and often do, move somewhere the tax bill is smaller. It also assumes the redistributed money reaches people in time and in a form that actually offsets what they’ve lost, which safety nets have historically been slow and imperfect at doing.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
It's certainly humble of the author to suggest an idea and then immediately criticise it. While I agree that Black Hat is a very important hat, I would also say let's not forget that there are other hats.
Black-HatThe black hat identifies risk. It is used for critical judgment and must give the logical reasons for concerns. It is one of the most powerful of the Six Thinking Hats.Credited to Edward de Bono
Read about it at debono.com ↗More on SignalCodes →
But it depends entirely on the state successfully taxing gains that increasingly belong to owners of capital who can, and often do, move somewhere the tax bill is smaller. It also assumes the redistributed money reaches people in time and in a form that actually offsets what they’ve lost, which safety nets have historically been slow and imperfect at doing.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
Focusing again on her second idea, the author states it simply: Everyone works less hours, not for less money.
Many people would like the sound of that.
Maybe it’s simply: everyone works less, for the same pay, because the point of productivity gains was always supposed to be that we’d eventually need to work less to get the same result.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
Again, however, as soon as she explains the idea, she goes on to explain why it won't work.
Mandating shorter hours at full pay is itself an act of redistribution, and it requires the same kind of enforcement power a robot tax does: labour law with teeth, a state willing to defend it, and employers who can’t simply route around it. And they can route around it. Faced with a mandatory shorter week, an employer can cut headcount instead of hours, push automation harder into the hours that remain, or relocate the work somewhere the mandate doesn’t reach. That’s not a different problem from tax flight. It’s the same problem in different clothes.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
And again I say: Yes, I agree - a lot of time and effort has been saved and misery avoided by pointing out flaws in new ideas, but too much criticism is going to block progress.
Black-HatThe black hat identifies risk. It is used for critical judgment and must give the logical reasons for concerns. It is one of the most powerful of the Six Thinking Hats.Credited to Edward de BonoRead about it at debono.com ↗More on SignalCodes →
Mandating shorter hours at full pay is itself an act of redistribution, and it requires the same kind of enforcement power a robot tax does: labour law with teeth, a state willing to defend it, and employers who can’t simply route around it. And they can route around it. Faced with a mandatory shorter week, an employer can cut headcount instead of hours, push automation harder into the hours that remain, or relocate the work somewhere the mandate doesn’t reach. That’s not a different problem from tax flight. It’s the same problem in different clothes.
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
Nearing the end of the article, Ms Jomard floats the idea of trying to implement both proposals, but dismisses it suggesting the government would run out of ammo fighting on two fronts simultaneously.
So I don’t think “build both” is the clean resolution I first reached for. A government or a labour movement has finite capital - political, fiscal, organisational - and winning a robot tax and winning a mandated shorter week are both expensive fights that may draw on the same finite reserve of leverage. Presenting them as compatible, additive policies skips over whether a society can actually afford to fight for both at once,
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
In her final paragraphs, Ms Jomard seems to have convinced herself that nothing is going to work.
The time route depends on a form of leverage - labour’s ability to make withholding its work costly - that AI is built to erode. If that’s right, the dividend most within reach may not be the one that feels most humane. If AI erodes the leverage that won shorter working weeks last time, what replaces it -and is there a version of that leverage a labour market weakened by AI could still exercise?
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
We can't even go on strike anymore!
The time route depends on a form of leverage - labour’s ability to make withholding its work costly - that AI is built to erode. If that’s right, the dividend most within reach may not be the one that feels most humane. If AI erodes the leverage that won shorter working weeks last time, what replaces it -and is there a version of that leverage a labour market weakened by AI could still exercise?
— Asa Jomard
The AI Dividend: Taxed, or Spent as Time?
So what do I think?
So overall, what do I think?
I think Asa's post tries to address an important question that certainly needs to be addressed: ie. How will people earn money when there is widespread job loss?
I worry, however, that she was too concerned about having perfect ideas and rushed to point out flaws rather than seek out possibilities and benefits.
Here is the concept of 'po'. If you follow the link, you see that Edward de Bono anticipated way back in 1972 that computers would take over a lot of human thinking, which of course we are now seeing happen through AI.
He also suggested that when computers take over human thinking, po will become the most important word for humans to use.
poProvocation, hypothesis, possibilities. Suspend judgement while you explore for new concepts and values.Credited to Edward de Bono
Read about it at debono.com ↗More on SignalCodes →
In the context of the present discussion, rather than float fully-formed ideas, Asa might have presented the idea as a po:
- po everyone works less, for the same pay
All sorts of questions might flow from that:
- Which people want to work less and which do not?
- Are there reasons we work the hours we do? What are they?
- The po suggests that workers should get the benefit of AI productivity. Should they be the only beneficiaries? Who else should benefit?
- Same pay suggests stability, and certainly many people are used to getting a stable income. How can we ensure income stability for those who need it? Can people be trained to cope with fluctuating incomes?
Has she used po, Ms Jomard's essay might have ended up as just a playful exploration.
Given how serious and fundamental these issues are - a playful exploration might be just what is called for.
Final comments on Signal Codes
So now let me briefly review what the Signal Codes had me do and what the consequences were.
- I had to nominate my source and provide you an easy way for you to get to it.
- I had to quote directly from the source allowing you to see what I included and what I left out.
- I knew that you would be able to click on any quote to see it in context - reminding me me again that I had to portray the author honestly.
- My interpretations of Ms Jomard's work were placed side-by-side with her own words. Very easy for you to make a judgement about whether I'd simplified or oversimplified what she had to say.
- When it came time to represent my views, I was encouraged to do so using concepts - forcing me to be both concise and precise about what I meant.
In summary, I was forced to behave myself. I had to respect both the original author and my reader as fellow human beings rather than as mere stepping stones to tread on in pursuit of my own agenda.
No doubt some people will rail against the whole idea of everyone developing ideas and appreciating ideas in others. They will continue to drift passively, guided by succession of self-styled luminaries who will serenade them with sweet sounds.
To those people I say: Leaving thinking to others has got society into a lot of trouble. It might be time to try something else.
Sources
The AI Dividend: Taxed, or Spent as Time?