![](/_next/image?url=https%3A%2F%2Fstorage.googleapis.com%2Fmantic-markets.appspot.com%2Fcontract-images%2FErnie%252F4feac0e3979b.jpg&w=3840&q=75)
Midjourney will be able to create images of people with different facial expressions by end 2024
Mini
9
Ṁ2212025
67%
chance
1D
1W
1M
ALL
On Dec 31 2024 I will run this prompt in Midjourney five times with random seeds: "a portrait of a smiling man and his frowning, sad wife"
I will run it with all the default options, aspect ratio 3:2, and the most advanced general model a paid subscriber can get under 100$ a month. If at least 50% of the resulting images successfully make the man look happy and the wife look unhappy, the claim resolves True.
Today the success rate is 0%. Five images generated with these settings are inlined below. Currently, Midjourney doesn't track individual subjects clearly in this way.
Image from 12/20/2023, still not able to distinguish the people's facial expressions.
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2FKeNekeZfmK.png?alt=media&token=a4f28b5a-665f-4ef5-ad54-527356316cd5)
Initial set of images from 5/10/2023 here:
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2F5JvfdroPGo.png?alt=media&token=0ef71380-798e-406f-b3b0-ba50bafea45b)
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2FAaDpY74P50.png?alt=media&token=2fadb957-2852-45de-b131-ac4f698a4da5)
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2Fnoeo_Dk9HY.png?alt=media&token=3ea3da1b-cc25-4070-881f-0c4360190d82)
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2Fo-24tzzTWB.png?alt=media&token=9e8effb3-6a96-4e62-b0a8-416ba9ec6921)
![](https://firebasestorage.googleapis.com/v0/b/mantic-markets.appspot.com/o/user-images%2Fdefault%2FbSQquI8VAN.png?alt=media&token=0c19a350-7891-4507-acd9-3e8246a5d9fc)
Get Ṁ600 play money
Related questions
Instant deepfakes of anyone (doing anything) by the end of 2027?
80% chance
Midjourney can draw a simple stack of colored shapes by mid 2024
45% chance
Twitter + Ai Image Generation + Bitmoji integration by end 2027
45% chance
Midjourney images can be effectively watermarked by 5/2025
22% chance
Will MidJourney add support for in-house AI Video Generation in 2024?
80% chance
The default number of images produced by one "/imagine" command in Midjourney increases to more than 4 by end 2024
61% chance
When will Midjourney first be able to create correct text in images? V6 may do it!
Will AI image=> video be pwrfl enough to make a person in an image rotate/walk/change clothes by EOY 2025?
71% chance
Will Midjourney release a public API by the end of 2025?
80% chance
Will AI image=> video be pwrfl enough to make a person in an image rotate/walk/change clothes by EOY 2024?
83% chance