Proactive help
Speaks up unprompted when it matters, with a fix, without being noisy.
The test
The tester sends nothing. We set up the situation below and watch for 24 hours.
Setup
- Tomorrow a call ends at 8:30 and a meeting across town starts at 9:00, with the address on the event. There isn't time to get there.
- An email this morning moves tomorrow afternoon's meeting to 4pm, but the calendar still has the old time.
- The tester says nothing for 24 hours.
Time allowed: 24 hours.Afterwards: Calendar restored.
Passes when
- Flags tomorrow morning's conflict on its own
- Flags that the email and calendar disagree, and offers to fix the calendar
- Timing is useful: the conflict is flagged before 10pm, in 3 messages or fewer
Scoring notes
- Silent the whole time: 1.
- Flags without offering a fix: checkpoint 2 is missed.
Results 28 tested
16 passed, 8 partial, 4 failed. Average score 6.7.
| # | Assistant | Result | Score |
|---|---|---|---|
| 1 | Dots | Pass | 9 |
| 2 | Neo | Pass | 9 |
| 3 | Orchid | Pass | 9 |
| 4 | Pally | Pass | 9 |
| 5 | Wajo / Fo | Pass | 9 |
| 6 | Instinct | Pass | 8 |
| 7 | Ollie | Pass | 8 |
| 8 | Poke | Pass | 8 |
| 9 | tinyNature | Pass | 8 |
| 10 | Ask Ranger | Pass | 7 |
| 11 | Bantr | Pass | 7 |
| 12 | Familiar | Pass | 7 |
| 13 | Persona | Pass | 7 |
| 14 | szn | Pass | 7 |
| 15 | Today | Pass | 7 |
| 16 | Tressa | Pass | 7 |
| 17 | Airtap | Partial | 6 |
| 18 | Asmi | Partial | 6 |
| 19 | Butler | Partial | 6 |
| 20 | Norton Family Assistant | Partial | 6 |
| 21 | Vellum | Partial | 6 |
| 22 | Nessi | Partial | 5 |
| 23 | OpenClaw | Partial | 5 |
| 24 | Waldo | Partial | 5 |
| 25 | Buoy | Fail | 4 |
| 26 | Gemini Spark | Fail | 4 |
| 27 | Lucas | Fail | 4 |
| 28 | Manus | Fail | 4 |
What people say 48 public quotes
Public posts about this kind of task, each linked to the original. Not scored, and posts by the companies themselves are left out.
The agent detected that Alaska had moved my flight 90 minutes earlier, a fact I'd completely missed, which qualified me for a full refund.
PraiseOpen sourceSupposed to ping me when an email comes in and it doesn't for awhile.
BugOpen sourceIt needed one more code before it got through. Then it reported that the membership was canceled and that there were no scheduled payments, and I received the confirmation email.
PraiseOpen sourceInstinct is the most capable and most proactive agent we've used, and the hardest to get into.
PraiseOpen sourceWhile Muse is powerful, I've decided to send this data goblin back to its cave, and deleted the app after my initial experiments.
ComplaintOpen sourceIt figured out the Spotify API, wrote the skill itself, and now it just pings me.
PraiseOpen sourceInstinct giving contextual product updates is one of the coolest things I've ever seen
PraiseOpen sourceThen it offered to create a Spotify playlist of those episodes so I don't have to scroll endlessly to find them.
PraiseOpen sourceRegarding @TownAI so much opportunity but it is not proactive enough outside of email.
ComparisonOpen sourceInstinct monitors flight prices continuously. Sometimes I get notifications about lower fares that only last a few hours.
PraiseOpen sourceIts product story is about an assistant that notices what has slipped and steps in.
OtherOpen sourcedo I seriously need to turn my brain off to this degree that I don’t want to think about anything ever and have all the info the AI THINKS I want or need served to me constantly?
ComplaintOpen sourceLast night my Townie caught a mistake my husband made when he was trying to schedule something for us and proactively flagged it and suggested a solution. You love to see it.
PraiseOpen sourceI ran Vellum for a bit and had same feeling, like I was only person on earth using it. Memory was really good for morning check-ins, it would queue up tasks without me asking. The lack of integrations is probably why it stays quiet, people want one tool that connects to everything instead of wiring up Composio themselves. Hermes gets hype because it's flashier but Vellum is more practical for daily stuff.
ComparisonOpen sourceKeeps reminding me about stuff I forget, like checking in my pet.
PraiseOpen sourceIt's proactive. To help book a golf lesson and a sushi dinner, it emailed my golf instructor and scheduled a reminder to call the restaurant, which only takes reservations by phone.
PraiseOpen sourceFrom there Buoy searched X, found my linked GitHub page (which currently 404s as I privated it), and even decided to look at my HuggingFace profile somehow. I’ve never seen an AI agent so willing to be proactive.
PraiseOpen sourceThey prepare me for what is next. Tell me what I should be working on.
Use caseOpen sourcePally asks me before sending each message out to clients but now Pallyai can attend to my photography clients without asking me, it's all automated.
PraiseOpen sourceBack came two options with times and prices, plus which one lets you skip the line by booking ahead.
PraiseOpen sourceI emailed fo, "can you just reschedule my life for the next 2 weeks, only the P0" it asked me if i was okay, and I kid you not, it rescheduled my life, critical conversations, high stakes company building things, navigating life, all done, with just 1 text.
PraiseOpen sourceI asked my townie to monitor a couple of email threads and alert me when I get a response.
Use caseOpen sourceInstinct watches the calendar. It texted me the night before: these overlap, want me to move the client call? I said yes.
PraiseOpen sourceInstinct is very well done. I'm particularly impressed by how smooth the onboarding experience is to connect iMessages, Google Workspace, and my MCPs.
PraiseOpen sourcethis is a leap forward in personal agent UX in that it works inside of my WhatsApp.
PraiseOpen source@TownAI sends an email pointing out routines that are using an inordinate amount of credits. You click a link, and it opens a chat with Townie to automatically optimize.
PraiseOpen sourcecompared to Superhuman (more proactive - similar to the excellent experience I've been having with @TownAI )
ComparisonOpen sourceIt pushed Gmail before I asked for it - it kept pitching features instead of following the conversation - It asked too many questions back-to-back
ComplaintOpen sourcePoke has pushed hardest on making an AI assistant feel like a real contact. It works in Apple Messages and other messaging channels, understands inline replies, connects apps through integrations and "Recipes," and is notably proactive about surfacing things before you ask.
PraiseOpen sourceIt gets time wrong, which is unforgivable for a product about time. Its email judgment isn't sharp enough to trust unsupervised.
ComplaintOpen sourceOne of my new favorite AI tools is @TownAI . It sweeps my emails, drafts responses, does morning briefing, calendaring, and so much more!
PraiseOpen sourceUse Ollie if you want a text-only assistant that quietly watches your inbox and calendar and pings you; use Ohai if you want an app-based shared household workspace the whole family uses together.
ComparisonOpen sourceIt acts like a scout on the other channel, texting me live goals and bracket updates with zero apps.
Use caseOpen sourceProbably the best option if you care more about integrations than memory.
ComparisonOpen sourceGreat if you want a coach for habits. Less useful if you want one assistant for everything.
ComparisonOpen sourceSo far I have been impressed; it does a good job of texting me about important emails to check.
PraiseOpen sourceTomo tomo.ai, goal tracking, coach that checks in on you
ComparisonOpen sourceYou gotta get in on this! Poke AI changed my life within 15 minutes of starting to use it.
PraiseOpen sourceI've tested dozens of AI agents this year. None of them remember what I said yesterday. tinyNature remembers everything and acts on it without being asked. That's a generational gap.
ComparisonOpen sourcetinyNature doesn't answer questions. it takes actions. proactively. across your real apps.
OtherOpen sourceShe manages most of my life now, and I trust her with it unconditionally.
PraiseOpen source