Firefly can't count

I am writing prompt for Seven (7) running horses but it's showing 5, 6 but not exact 7 horses.

When having questions about an AI, ask an AI... this is from a long chat I had with Google’s Bard:

Me: in my experience, AI has problems counting, both when chatting and when generating images. Why do you think this is?

Bard: You're absolutely right! AI image generators do struggle with counting in images compared to humans. There are a few key reasons for this:

• Data-driven learning: AI models learn from the data they're trained on. While they have huge datasets, representing all possible variations of counting accurately can be difficult. For example, if an AI sees mostly images with 2 apples together, it might struggle to differentiate 3 apples later.
• Lack of abstract understanding: Humans understand numbers conceptually, but many AI models don't. They see objects as individual shapes and colors, not abstract quantities. This makes it challenging for them to accurately count across different arrangements and contexts.
• Focus on image generation, not object recognition: Some AI image generators might prioritize creating an aesthetically pleasing image over ensuring precise counting. While they might capture the overall scene well, the exact number of objects might be inaccurate.

Here are some additional factors that can contribute to counting issues:

• Object variations: Size, position, overlapping objects, and other variations can confuse the AI, making it hard to distinguish individual instances.
  Background complexity: Cluttered backgrounds can further complicate the process of isolating and counting objects.
  Limited training data: As mentioned earlier, more diverse and specific training data could improve counting accuracy.
• Researchers are actively working on improving AI's counting abilities. Techniques like incorporating object recognition modules and using synthetic data with controlled variations are being explored.

Overall, while AI has made significant progress in image generation, counting accurately remains a challenge. Understanding the limitations and reasons behind them is crucial for interpreting AI-generated images and using them appropriately.

So, the correct answer to “Why can’t Firefly count?” is “Because image generation doesn’t work like that.”

..."I am not sure, @CMass , who marked this as the "correct answer", but it is not."

Come on, droopydog500, you can't just leave us hanging... you must know... what is the correct answer? :))

If you prefer in simple terms. That would be far too expensie and require yet years to gather data and completly new model to include such simple for humans and yet difficult for AI model thing to do. I would estimate a milions of budget and at least 3 years of reserach.

I am not sure, @CMass , who marked this as the "correct answer", but it is not.

Just a quick note, Maciej...

You mentioned that it's very rare for people to count items when describing images in text. Do you mean when they are experimenting with firefly or when they are using AI for real world projects?

I'd like to point out that the people in this thread who posted examples are those "rare" occurrances you refer to. In other words, I personally don't think it's rare (in my line of work) to ask for precise numbers when creating imagery. I do it constantly.

My client wants a beer ad with three bottles. Classic configuration.

The football coach wants a poster with eleven players (a full team) to inspire his JV team.

The baseball coach wants a poster with nine players (a full team) to do the same.

The dog breeder wants an ad with four dogs.

In my world of photography, graphic design and commercial art, specific quantities are normal.

(for what it's worth, I'm not trying to challenge your assertion. I'm merely curious about your perspective and want to know more because it doesn't fit my experiences.)

If you prefer in simple terms. That would be far too expensie and require yet years to gather data and completly new model to include such simple for humans and yet difficult for AI model thing to do. I would estimate a milions of budget and at least 3 years of reserach.

I appreciate everyone answering my question about AI not being able to count. And I appreciate the explanations of how developers didn't include counting in AI training. It helps me understand that AI wasn't programmed to include counting. And, I suppose if you understand how/why AI works, you understand why counting wasn't included in the training.

For what it's worth, I have been trying to clarify your statements to make sure I understand you. That's why I used phrases like, "it sounds like you're saying". So, when we say counting wasn't in the training, it sounds to me like the trainers didn't consider it "of significance or consequence" to include it at that time. Or... maybe it was too difficult to include. Or maybe too expensive. Either way, the end result is the same. Counting wasn't included.

And I thank everyone for helping explain that.

However... and you knew this was coming... I still don't understand "why" it wasn't included. And if you don't know "why" either, that's cool. We're in the same boat.

Counting is of base-level importance to human development, so my human brain naturally assumes it would be important to an artificial intelligence designed to help humans (assuming that's what AI is here to do). That's where I am right now. I don't get it. I get that it wasn't included in the training, but I don't get why it wasn't included.

And that may not be a question anyone here can answer, since we're not AI developers.

I'm cool with that. Maybe one day an AI developer will help shed some light. 🙂

It is very easy answer to why AI that create images can not count.

That was aparently not in the trained set of data. When people describe the images in text, it is very rare that they count number of items. We usually use other quantitative words like a lot, a few, a croud, etc. Therefore without a special training for that particualr use case the AI will never be able to count or understand that less precise words for quantity.

- Independent software developer and artist.

It sounds like you're saying the developers of AI don't consider counting to be important.

By @holtwebbart

I couldn't find where either of us used that word, but I found 10 places where you used it. As for me, I don't think it. All I know is counting is not there (yet).

• Both of us are volunteers who don't work for Adobe. 
• We are not developers.
• Counting is not working in any generative AI. It's not just Firefly.
• The hope is that it will work someday and that it's being worked on.
• Developers rarely share want they are working on. They wait until they have something that works.

I hope this conversation has helped. If not, I hope someone else will step in. I can't think of anything else to add without repeating myself.

Jane

Well, creating from scratch when AI doesn't work takes us outside the scope of the AI discussion. That's just doing things "old school". 😄
Good article. Thanks for digging that out. It talks about how AI can’t make out letters or numbers because it’s trained on image data only. Would it make sense to incorporate OCR (optical character recognition) into the AI platforms? Technology has had the ability to recognize text and numbers for decades.

AI can accurately reproduce a life-like animal in a variety of realistic scenes but has no idea how many animals there are in that scene. I wonder why the two can’t work together. AI should be able to use OCR software.

I understand that it doesn’t work that way, but I don’t understand why. Numbers and letters are so integral to human development that I’m baffled why the creators never addressed it.