Does an AI scribe put in the orders?

Mostly no. It writes down what you said about the labs. It does not send them. A nurse practitioner asked that question in a thread this year and nobody answered her. She had a human scribe once who placed the orders. Now she does that part herself, at the end of the day, from memory and a note.

What did the human scribe actually do?

More than type.

The nurse practitioner who asked was comparing two things. A person who sat in the room, heard her say she would send the panel, and sent it. And a tool that writes a paragraph saying the panel was discussed.

Both produce a record. Only one produces the thing that was promised.

That is the gap the word scribe hides. A human scribe was a pair of hands. Most of these tools are a pair of ears.

Why is an order different from a sentence?

Because a sentence can be nearly right and cost you nothing.

If a draft says three days of symptoms and it was four, you fix it in two seconds and nothing else follows. If the wrong order goes in, something happens out in the world, and it happens quietly.

One nurse practitioner said the tool built into her records system picks the wrong things and uses the wrong diagnosis codes. She uses it to remember what she discussed, and she still places every lab herself. That is a sensible way to work. It is also an admission that the automation stopped at the door of the only part with consequence.

What happens to the promise you made in the room?

It goes on a list in your head.

The patient heard you say it. They are waiting for it. Between that moment and the end of your day there are fifteen more patients, and the note, whatever wrote it, does not chase you.

The one clinician in that thread who sounded genuinely relieved was not relieved about the prose. She said she no longer forgets promised orders when she charts at the end of the day. Same category of tool, different job. Not a better sentence. A shorter gap between what you said and what got sent.

Can you check it in thirty seconds?

Yes, and the habit is worth building whatever you use.

Say the orders out loud in the room, so whatever is listening has them. At the end of the visit read only that part of the draft, not the whole note. Compare it with what you actually sent.

Thirty seconds, one section, before you sign. If the tool cannot show you that list on its own, you have learned something about the tool.

What should you ask before you buy one?

One question, and it is not about accuracy.

Ask what happens to an order the note says you promised. Does the product show it to you as a list, hand it over as a draft for you to approve, or write a sentence about it and move on?

Most will answer with the third. That is not a scandal. Placing orders safely is a much harder problem than writing a paragraph, and the right answer today may well be that a person does it. But you should know which of the three you are buying, because the marketing for all three sounds the same.

The short version

Mostly no. AI scribes write down what you promised and leave the sending to you, so the last mile stays manual. Solanto will not claim otherwise until it is real. What it is being built to do is keep that list in front of you instead of burying it in a paragraph, in Cloud mode or in Private Mode, where the note never leaves your device.

Solanto has not shipped yet. The price is on the page rather than behind a call, and early access is open: solanto.ai/pricing

Ask the colleague who loves their scribe one question. When did it last remind you of something you promised and did not send?