Artsmart Avis 3

Excellent image generation tools with good realistic backgrounds. User friendly and fast delivery.

Ai art is fast becoming a part of every small business arsenal. However, not every application has a fun, accessible and easy-to-use UI. ArtSmart does! Great pics, great customer service, great product.

I was looking for sustainable alternative to MidJourney.
Although AS uses Stable Diffusion, the quality and style of images produced are easily comparable for me.

Before I even used AS, I tested their customer service. Josh handled my queries intelligently, with an engaging sense of humour and within reasonable time.

Workflow window

The way MJ does it, displaying version numbers doesn't seem useful as the older ones aren't going to be usable long-term.

Whereas, AS (playground) simplifies things, I only have to pick between 3 generators; Standard, Semi-Real or Hyper-Real (in beta). This is far more useful. Additionally, unlike MJ, AS doesn't eat up more credits just because a generator is was updated.

Whilst this tool is totally fine in generating fantasy/supernatural creatures/environments,
I think Realism is where the long term value is going to be. Glow-ups, Beauticians and Spiritual Manifestors are going to capitalise on this for sure. Ever seen something beautiful after the trend had already passed?
I think these AI tools will help us create our own trends, or kickstart our imagination at the very least.

For now, I've already seen some incredibly lifelike humans generated by Semi-Real and Hyper-Real.
I almost forgot I wasn't viewing a dating profile photo.
Still, I reverse-google imaged them.
These people really don't exist in IRL lol.

My Four long-term Value predictions

1. For those of us who have experienced chronic loneliness/ low moods/illnesses,etc, I think coupled with a visual avatar tune like Pipio, AI image tools like AS is going to help those of us improve our emotional wellbeing and confidence with real-life interactions.

2. But I believe the future of these Image generation tools comes down to how well we can iterate.
Prompt scripts are great but they're only perfected after a lot of experimentation.
Whilst their image to command tool is a great catalyst, a lot of the time, an otherwise near-perfect image is ruined by the a doppelgänger, weird face or the fact that I changed my mind and don't want the subject to wear glasses.
I've spent a lot of time tinkering in Photoshop, so I plan to do the same with inpainting and repairing weird faces (which seems to improve the more times I run the fix).

3. The main appeal of AS to me is the potential to replace GauGan (aka NVIDIA CANVAS).
I use it to paint the environment based on the detailed image in my head. But I can't command it as well as AS. But further down the road, when Outpainting is released, I believe AS could become a one stop shop for me.

4. Beyond that, I think Replicate/GIF-like dynamic images is going to attract a whole new customer-base.

I'd like to see the minor technical annoyances get resolved (like the doppelgangers, processes getting stuck). Like other platforms, AS are trying to provide fixes for Stable Diffusion's shortfalls. And there's always the Discord community for problems that simply need a command tweak (for example, when the 'no' prompt templates don't work).
Also, I'd like to see free preview to help protect our credits from being depleted when we're simply demo-ing/sketching a prompt. Most of us will be trying to import prompts designed by others or converted from MJ. And there's really no way to guess how well it's going to translate until we try it.

Finally, for those who are worried about ethics after Mid Journey got challenged, watch Youtuber LegalEagle's 'AI Versus The Law' video from 16th minute onwards. I hope that puts things into perspective for you as it did for me.