This may be my first negative AppSumo review, and I've tried more than my fair share of products, but I couldn't ignore this one. Mostly because it is one of the few *non-refundable* products I have seen that I thought would be worthwhile. While I knew the AI rendering solution Stable Diffusion was freely available and easily implemented, I saw real value in SuperMachine saving me the time of doing things myself. How difficult could it be to create a UX to run commands against an existing solution and receive the output? Apparently, it isn't as easy as it sounds, and it seems that someone else came to the same conclusion during its development. I guess a little hubris can lead to a lot of trouble.

There is also a pretty poor value proposition here. As mentioned by others, the credit system and how it's presented to the end-user, outside of purchasing the product, is dubious. It's like trusting a mechanic with a blank check when you take a car in for a checkup. While you may see 750 credits a month, it's not a 1:1 cost and you don't realize this till after your purchase - a deceptive business tactic. The system by default will use 1 credit per generation, however, if you deviate from that setting then you can, at the time of this review, be charged upwards of 1.5 credits for an image. As a result, the value of the 750 credits has now drastically been reduced. Additionally, you are trusting that the developer will not inflate the cost of generations down the road, which brings me back to the dubious implementation of the credit system and trusting the business mentality going forward.

Moreover, I experienced issues where my prompts wouldn't produce anything, but still would consume credits. Approximately 25%-33% of the prompts that I generated ended up not having outputs and there was no way to find out why -- nothing in the prompt was inappropriate and regenerating the image again with the same prompt would render an image. So this was just another area where my credits were being further devalued.

The other BIG issue is that when you do get an image that works well it's pretty much the size of a thumbnail and isn't really useful for most usecases today. Sure, I can spend time (and money) upscaling it and get a larger, but degraded, image. But why isn't this part of the system to begin with? Because this is a very early stage product that feels rushed. I would go as far as saying that it really doesn't belong on AppSumo.

AI art is hot right now and this is a poor implementation of what is out there. There are a lot of developers focusing on it because it's easy to implement. So if you are going to buy something, buy something that works for your usecase and buy something that has been trained for that usecase. Midjourney and Dall*E do amazing jobs of generating images because they trained those systems well. SuperMachine, ironically, has trained me to be more vigilant in questioning products that have a no-return policy.