LLMs answer the sentence. Manwe answers the situation.
Same frontier models, six role-bound advisors, multiple rounds of structured debate. A reframe and a grounded verdict instead of a fluent guess.
Soft launch: the Manwe web app is live today.
Start one decision: https://t.co/fN5WFl5Xmz
Manwe v0.9.5 is live.
This release adds Diagnostic Trials.
Instead of asking one model for one private answer, Manwe can now put a question on trial: multiple model-backed teams, evidence, debate, cross-exam, audit pressure, and one independent verdict.
Download for free:
https://t.co/zdf4aS7p8v
@dominguezdaniel Try with Manwe’s Mac app https://t.co/32LlWf01p3
I have a few users who claim that Manwe is very accurate with football predictions if provided with enough data.
I mean, for months they ran the doom marketing and made sure that everyone knows ‘how dangerous Mythos is’.
Then they released it anyway claiming that it has security gates which then bypassed by two engineers in Amazon.
I loved Fable while it lasted but Anthropic self-inflicted this to themselves.
@Tancrededib You have an idea but don’t have anyone around you who can challenge with your assumptions. And you also don’t trust AI answers as ‘you are absolutely right’.
Then you ask Manwe: https://t.co/32LlWf01p3
@DanielSmidstrup I have one-time purchase little apps that make $1k monthly 🤷♂️
Now another one is coming but it’s not little anymore but a proper video editor (with on-device AI of course).
@thomvlieshout I’m definitely working more than ever but at least finally enjoying the work itself after a lot of years.
I was so bored from whole ‘design the interface and hand it over and pray everything goes okay’ process.
AI makes me enjoy being a designer again -and build wild things-.
@tibo_maker@pbteja1998 He doesn’t say ‘he likes GPT’. He says ‘he likes the Codex app’ which I agree. Only blocker for me to switch to Anthropic is their horrendous apps.