In the AI Era, What Remains for Humans Are 'Decisions' and 'Responsibility'

I previously wrote an article titled 'Asking Questions Has Become My Job'.

Since I started using AI frequently, tasks like researching, organizing, and comparing have become overwhelmingly faster than before. Compared to when I used to browse websites and official information on my own, I feel that the very way I work has changed.

So, after AI starts providing answers to our questions, what is left for humans?

For now, I believe it will converge into two main things.

'Making decisions' and 'taking responsibility for the results'.

Decision-making is not about choosing the correct answer.

If you ask AI a question, it will provide multiple options. It will organize the pros and cons of each, and if you provide the conditions, it will even suggest the best option.

However, you must decide for yourself which option to adopt in the end.

In my work, I make small decisions every single day.

How much research is enough? Can I trust this information? Should I publish this text as is? Should I prioritize accuracy and delay publication? Should I go for immediate reactions or protect long-term trust?

There is no single correct answer that applies to all of these judgments.

Choosing something also means giving up something else. If you choose speed, the time spent on verification decreases. If you increase the amount of information, readability may be lost.

AI can help up to the point of laying out the options. However, it cannot decide what to prioritize and what to give up.

I believe that decision-making is not the task of guessing the correct answer, but the work of setting your own priorities.

Responsibility means not severing your connection to the final output.

The other thing that remains for humans is 'taking responsibility'.

There was a mistake in the text created using AI. The result of adopting an AI suggestion did not go well. My judgment caused trouble for someone.

Even if I say 'because the AI answered that way' at such a time, it does not mean my professional responsibility disappears.

Using AI and making AI responsible for your judgments are two different things.

Admittedly, when we talk about taking responsibility, it is easy to imagine only receiving punishment or apologizing. However, I think responsibility in individual work is a bit broader than that.

Correct mistakes if they occur.
Explain why that decision was made.
Face those who were inconvenienced.
Change the approach to ensure the same failure does not happen again.

In other words, isn't responsibility about not severing the connection between yourself and the work or decisions you produce?

The more capable AI becomes, the heavier decisions weigh.

As AI performance improves, human work will become easier. This is certain.

On the other hand, there is also a danger that the process of thinking will become less visible.

When plausible answers are displayed immediately, it is easy to decide 'this is fine' without thinking deeply about whether to adopt them.

I myself have come to work under the assumption that high-performance AI exists. I also feel anxious about whether I would have the ability to think from scratch if it suddenly became unavailable.

The more accustomed we become to convenience, the more we want to entrust not just questions, but also judgments, to AI.

However, as AI becomes more capable, it does not mean human decisions become unnecessary. Rather, I believe that because the time spent on research and organization has been shortened, the weight of the final decision—what to choose—increases.

Human work remains before and after AI.

If we simplify the workflow, it looks like this.

First, humans decide what to ask.
Next, AI provides information or options.
Then, humans decide what to adopt.
Finally, humans take responsibility for the results.

The scope covered by AI will likely continue to expand.

Even so, human work will remain before and after AI.

What to ask.
What to believe.
What to choose.
How to face the results.

I suspect that rather than human work disappearing, the core of our work is narrowing down to these points.

Deciding rather than providing answers.
Taking responsibility for the results rather than just finishing tasks.

Work in the AI era might unexpectedly converge on these two things.

This time, I thought about this from the perspective of one person working using AI. However, for those in positions where their decisions affect many people, the meaning and weight of responsibility change.

I would like to consider the responsibility of those in 'decision-making positions,' such as business leaders and politicians, in the next article.

▼Read also

・The more you use AI, the more you need a 'record of judgment'

https://note.com/poliplus/n/n0619b141a870

・AI operational design for deploying a single video across multiple media—decisions and responsibilities left to humans

https://note.com/poliplus/n/nfa84bc37af7a

▼Summary of this theme

・AI x Communication know-how summary

https://note.com/poliplus/n/ncd6194c20727