I put Muse, Dot, and Instinct to the test. There was a clear winner, but not because it worked best.

I've been playing around alongside assorted AI agents — Muse, Instinct, and OpenAI's Dot, and I've established them genuinely really helpful for certain things.

But which one is best?

I put them to some, uh, semi-rigorous testing. Ok, not totally rigorous or controlled, but I gave them all the identical 3 tasks — triage my email, command me lunch, and build me a individual website — and it was engaging to see which worked best. It was additionally engaging to see how alternatively they functioned at these tasks.

If you haven't started using an AI delegate yet but are curious, well, I'm not certain that this is going to answer all your questions, but perchance is a location to get started.

A big spoiler current for my supreme recommendation: correct now, Meta's Muse is the lone liberated among the bunch, so in my mind, that wins for most normal group who fair desire to try an delegate out. OpenAI's Dot is lone for the $100/month scheme (ouch!) and Instinct is currently invite-only.

To me, that's a no brainer. Do the liberated one (unless you're already paying for the ChatGPT plan, in which case you're likely already a beautiful pro person and likely don't need my advice).

Challenge #1: Manage my email

I get a lot of email. Most of it I don't need to annoy opening. Help me fig out which are the crucial ones.

Muse: Muse has been a dutiful serviceman here, and this was beautiful much the archetypal project I gave it whenever I started using it a few weeks ago. So far, "triage my email" has been the one really helpful item an delegate has done for me.

Instinct: Overall, I don't akin the fact that Instinct plant chiefly through your content messages vs. a dedicated app. While that's extremely convenient in certain ways, I don't akin the idea of seeing email (bad, no fun) in my texts (good, fun) — mixing the two feels wrong.

However, I have to hand it to Instinct — it established a Paperless Post invitation for a gathering that I would alternatively have entirely missed in my individual email. It was the lone delegate to drag that one out for me.

Dot: Dot did item really helpful that required a lot of context, and it did all on its own. It noticed an older email toggle from final duration alongside my tax accountant (I had submitted for an extension, due in October). Dot noticed my accountant never confirmed he really submitted my taxes — yikes! This really really saved me! (Don't worry, the taxes had certainly been filed.)

Finding an older — but crucial — email toggle that I hadn't closed the iteration on is exactly what I desire from an AI agent/assistant. I awareness akin 90% of things I'd fairly do myself, but my individual brain gets effortlessly distracted and forgetful, and there are plentifulness of things I ignore to agreement alongside or get distracted.

Winner: Dot. Also me, for not going to jail for tax evasion.

Challenge #2: Order me lunch.

I asked all delegate to command me a dish at a particular local restaurant. Because I didn't desire to annoy genuine humans, I did the orders for pickup, to evade any disturbance or multiple distinct deliveries.

This turned out to be a genuine fiasco, and I learned my lesson: it would have been infinitely easier for me to fair do it myself.

Muse: Initially, Muse was extremely helpful. I accidentally gave it the incorrect eatery name, but Muse helped me discover the correct location and pulled up the accurate list options for me to see.

I had already connected my credit cards alongside Stripe, so I figured this would be easy. It offered up my debit cardstock or another Mastercard, which I assumed was my regular credit cardstock (I had fair gotten a new cardstock in the communication and didn't acknowledge the final four digits).

When I eventually went to inspect my wallet, I discovered it was a error — it was really the cardstock connected to my health insurance-related HSA account, which is lone approved for medicinal expenses.

By the period I straightened this disorder out, I had already moved on to placing orders alongside Instinct and Dot. When I went to choice up my orders at the restaurant, there were lone two bags alongside my name on them. Sure enough, I had gotten so distracted alongside the another orders that I never confirmed alongside Muse concerning placing the order, and it never sent through. (Can you see now what I was saying before concerning how I get distracted and don't complete tasks?)

Instinct: The most annoying part concerning Instinct, again, in my opinion is that it's complete content messages. That method there were several quirks whenever I tried to content on my laptop — sometimes it unsuccessful to dispatch as SMS. But following a few fails, Instinct was capable to location the command comparatively quickly. Well, kind of.

All told, this took 9 distinct content messages on my end — not including Instinct's responses. It is difficult for me to ideate that this was at all easier than fair going onto the restaurant's website or UberEats and doing it myself.

Although the procedure was fairly painless, there was one item I didn't like: it put the another through alongside no tip (!!!) and didn't ask me if I wanted to add one.

Dot: Somehow, this was the most difficult to get to work. It kept getting tripped up by the 6-digit code sext by content needed to login, but it kept failing (I had to get the code texted 3 distinct times), and the browser it created for me to act wouldn't let me category into it. A mess, basically, but eventually it worked.

Winner: None. I awareness extremely assured in saying that it would have been infinitely easier to fair location the command myself through regular means. It's imaginable that it would be easier to command in the future, since I've already done the activity of environment it up once. But I've learned my lesson.

Challenge #3: Build me a website

I have an old personal/professional website, but it's due for a refresh. I have awful imaginative awareness and don't desire to expend much period designing a site. Seems akin a fine job for AI to do for me!

Muse: I appreciated that, dissimilar Dot and Instinct, Muse didn't use the photo of my visage from my byline photo (it seems extremely weird to have a elephantine photo of your own visage on a expert website).

But all three basically did the identical thing: pulled the info from my author bio on Business Insider, selected several latest articles as clips, beautiful essential stuff.

Instinct: The scheme current was the most boring, and in a way I wouldn't equal cognize how to create it appearance improved (terrible imaginative sense!). Weirdly, the selected part it pulled for me were random ones from complete a twelvemonth ago.

Dot: The appearance and scheme were likely the finest here, and it pulled a decent mix of articles I'd desire to showcase. But GOOD GRIEF the tagline current is so corny that I'm kind of embarrassed to equal display you, equal although I didn't compose it myself:

Winner: Based on the mockups it made alongside scheme alone, I'd go alongside Dot. I think I could activity alongside this. But all three of these made specified cringey, cheesy copy concerning me that it honestly made me awareness genuinely depressed. Is this who I am? Yuck!

Obviously, I'd have to rewrite all the copy myself, and choice the articles for clips — and at this point, is this equal preservation me any period beyond being a extremely generic template? Ah well.

Conclusion: The biggest difficulty alongside all of them was figuring out what benevolent of project an delegate is well-suited for.

Most of all, I established that trying to oversee three distinct agents at the identical period was a huge hassle. I desire AI agents to be akin a wife — making my existence easier by anticipating my needs, not a hubby who merely does chores whenever told. But this has backfired on me: now I have three wives, and they're all nagging me and I'm forgetting which is which. Ultimately, I don't propose AI polygamy. May my test alongside it assistance you choose your bride.