What I mean by “the most useful thing” is, by definition, the most productive – that which will advance your goals the most for the least effort. Sure, costs in determining that matter. But it’s almost tautological that that’s what is best to work on.

But it’s not generally possible to determine what will actually advance your goals.

Imagine for a moment Queen Victoria had established an overarching goal, “produce a means of communication that will permit Us to send Our Voice to all corners of the empire”, and given it to her scientists and engineers.

How much progress would they have made?

Now, consider the equations produced by J.C. Maxwell. There were no indications ahead of time that his interest in electromagneticism would produce such a powerful result – a result that was instrumental in the eventual production of the radio, a device which fits Victoria’s requirements to the tee.

You can’t determine what the benefits of funding basic research will be. Yet doing so was the best way to reach the hypothetical goal, not funding research into the goal directly, which almost certainly would have produced a greatly inferior method within the bounds of then-known principles.

Aiming at the target is often the worst way to hit it.

I’m surprised the claim needs defending.

You’re not much of a rationalist, then. All claims need defending. Some can be defended with trivial effort, some require a great deal more – but ALL require it.

Welcome Aboard Patri!

A company will not make an investment/purchase if the expected return (%age) is less than the cost of capital. The cost of capital is determined by the rate of return that the money could get elsewhere WRT the risk.

In terms of a person doing something, the expected return is the easier part to calculate.

The risk is the hardest part to calculate by far.

I like Rubin’s advice “The Big Fact” in his Overcoming Indecisiveness; “In very few instances is one decision actually better than another.” His claim is that your commitment to making a decision work is usually more important than the particulars of the decision. Out and out bad choices are usually ruled out early in the decision process. I think this is one reason many people have trouble visualizing the situation with Eliezer’s paperclip maximizer, that is so obviously bad that it wouldn’t make it far enough in a human’s decision making process as to even be consciouslly considered.

Would optimum be more practical if we understand it not as ‘best’ but as ‘best that occurs to you’.

And does our action reveal what we really think is more important?

I’m surprised the claim needs defending. What I mean by “the most useful thing” is, by definition, the most productive – that which will advance your goals the most for the least effort. Sure, costs in determining that matter. But it’s almost tautological that that’s what is best to work on.

@Dagon – I am trying to make a simpler point. I often find myself doing things which “feel urgent” but are not long-term critical, and interfere with long-term critical tasks. Of course it is hard to determine the best strategy. But most of the time, I think most people don’t even try. They do what seizes their attention.

I’ve been thinking a little lately about the difference between doing something useful, and doing the most useful thing. The latter is a lot harder, yet a lot more productive.

Is it? Can you demonstrate that “the most useful thing” is much more productive than “doing something useful”? Can you even defend the claim?

One tidbit of ancient wisdom seems to be genuinely useful: concentrate on the problems you have here and now, keep your mind on what you’re doing and not what you’re going to do or could have done.

Never thinking ahead or about alternate possibilities is lethal. But we spend far more time and effort on such than we need or is good for us.