When None of Your Options Is Good Enough

September 9, 2026 · By Michel Valenzuela, Founder of Daryl

You have compared the options carefully.

One costs more than you can afford. Another misses a deadline you cannot move. The third fits the budget and timing, but fails to deliver what you actually need.

Which should you choose?

The question assumes that one of them deserves your commitment.

Sometimes the most useful conclusion is that the available options are inadequate. Recognising that gives you a different task: change the conditions, find another possibility, or decide whether the commitment should happen at all.

I think an AI assistant should make room for that conclusion.

The strongest option can still fall short

Ranking options tells you how they compare with one another.

It does not establish whether any of them meets your requirements.

Imagine choosing a supplier for an important project. You need delivery before a fixed date, a price within your budget, and a specific level of quality.

Supplier A offers the best quality but cannot meet the deadline. Supplier B can deliver on time but exceeds the budget. Supplier C is affordable and available, but its work would be unsuitable.

A comparison might place one supplier first. That ranking does not resolve the underlying problem.

Before asking which option is strongest, ask whether each is workable.

Separate requirements from preferences

This judgment depends on knowing what the decision must satisfy.

A requirement could be compatibility with an existing system. A preference could be a more attractive interface.

A requirement could be delivery before an event. A preference could be having an extra week for review.

If everything becomes mandatory, you may rule out reasonable choices. If genuine requirements become optional, you may accept something that cannot serve its purpose.

Write down the distinction.

For the supplier example, you might say:

“Delivery before the event is essential. The budget has a small amount of flexibility. The work must meet these three specific quality criteria.”

That creates a basis for evaluating feasibility before debating which option you like most.

Ask AI to check the options against the requirements

An assistant can help organise this examination.

Give it the available information and ask:

“Check whether each option satisfies the requirements I’ve listed. Identify any failures and anything we still need to confirm before treating an option as workable.”

That last part matters.

Missing information is different from a confirmed failure. A supplier whose delivery date is unknown needs a question. A supplier who has explicitly ruled out your deadline presents a different problem.

The assistant should keep those distinctions visible. You should also check that it has interpreted the requirements correctly.

A useful comparison might look like this:

Now you have a clearer picture of what prevents a decision and what could potentially resolve it.

Find out whether the conditions can change

An inadequate option may become workable through a specific change.

You could reduce the scope, divide the work into stages, extend a negotiable deadline, or increase a budget for a justified reason.

But each adjustment has consequences.

Reducing the scope may remove something essential. Increasing the budget may affect another commitment. Splitting the work may introduce coordination problems.

Ask:

“What would need to change for this option to become workable, and what would that change cost us?”

This helps you evaluate a revised arrangement rather than accepting a compromise without examining it.

Look beyond the original list

The options you have gathered may reflect the way you started searching.

Perhaps you looked only at suppliers you already knew. Perhaps you assumed the whole project needed one provider. Perhaps you treated a familiar approach as the only way to achieve the outcome.

Once the initial options fall short, revisit those assumptions.

Could the work be completed differently? Could part of it be handled internally? Is there another category of supplier? Does the project need to be this large?

An AI can suggest possibilities to investigate. Those suggestions still need verification, especially when they depend on availability, pricing, or capabilities the assistant has not confirmed.

The purpose is to widen the investigation in a useful direction.

Waiting also needs a reason

Declining to choose can protect you from an unsuitable commitment. It can also allow a deadline to pass.

If you defer, define what the delay is meant to accomplish.

You might need a revised quote, a delivery confirmation, or one additional candidate. Set a point at which you will reassess.

“We will request revised proposals by Thursday and decide on Friday using the same requirements.”

That turns waiting into a bounded next step.

If there is no time to gather more information, you may need to consider reducing the objective or cancelling the activity. Those choices deserve the same careful examination as the original options.

Make the trade-off explicit when you must proceed

Sometimes circumstances require action despite imperfect choices.

In that case, be clear about what you are accepting.

Which requirement will remain unmet? What consequence follows? Can you reduce the impact? Who needs to understand the change?

A decision becomes easier to review when its limitations are stated plainly.

Choosing under constraint is part of real work. The important thing is understanding the commitment you are making.

Why this matters for Daryl

Daryl is a personal AI for Mac, currently in development, designed around local conversation and support for decisions.

One of the behaviours I want to develop is recognising when the available choices do not justify a recommendation.

That means helping people distinguish between a workable option, an option needing more information, and an option that fails a genuine requirement. It also means making room to revise the problem or investigate another direction.

The examples in this article describe that design goal. They are not demonstrations of a released capability.

A useful assistant should help you understand whether a choice deserves your commitment.

Sometimes that leads to selecting an option.

Sometimes it leads to saying, “We need a better set of options before we proceed.”