A Balanced Perspective
Scrolling through Mastodon on a #Caturday, you encounter numerous cat images. While a few images include descriptive Alt Text, most lack this essential accessibility feature. Faced with this, what should you do? Here are several approaches, each with its own merits and drawbacks:
1. **Retoot with Hashtags AltForMe**:
Retooting with hashtags such as #AltForMe invites others to provide the needed descriptions. This approach leverages the community's goodwill and increases awareness about the importance of Alt Text. However, it relies on the altruism of others, which may not always yield timely or consistent results.
2. **Reply to the Tooter and Scold Them**:
By replying to the original poster and asking them to add Alt Text, you directly address the issue at its source. This method promotes education and accountability. However, it can be time consuming, perceived as confrontational, potentially discouraging positive responses, or creating conflict.
3. **Find a Friend to Describe the Image**:
Asking a friend to describe the image ensures you get a reliable and detailed description. This personal approach fosters social interaction and understanding. However, it can be inconvenient for both parties, especially if your friend is not immediately available.
4. **Sigh and Move On to the Next Cat Image**:
Accepting the lack of Alt Text and moving on is the path of least resistance. While it avoids conflict and inconvenience, it perpetuates the accessibility issue and leaves you without the context of the image, diminishing your enjoyment and engagement.
5. **Use an AI to Describe the Image**:
Sending the image to an AI for description offers a quick and often adequate solution. While AI-generated descriptions might not be perfectly accurate, they provide a broad understanding of the image's content. However, this approach raises questions about the long-term implications of relying on technology for accessibility.
### The Future of Computer Vision and Alt Text
Considering the potential advancements in computer vision, two critical questions arise:
1. **Will Computer Vision Ever Approach the Accuracy of a Friend? **
Technological advancements suggest that computer vision will continue to improve, eventually reaching or surpassing a level of accuracy comparable to human descriptions. However, while AI can process and interpret visual data rapidly, it may struggle with nuances, context, and emotional subtleties that humans effortlessly provide.
2. **If AI-generated Alt Text Becomes Prevalent, Will It Harm or Benefit Humanity? **
The widespread adoption of AI-generated Alt Text could bring significant benefits, such as making the internet more accessible at scale and reducing the burden on content creators. However, it might also lead to complacency, where people no longer see the need to add personalized Alt Text, potentially diminishing the richness and accuracy of descriptions. Moreover, over-reliance on technology could erode human skills in creating empathetic and contextually rich descriptions, impacting the quality of communication, and understanding.
### Conclusion
Navigating the challenges of Alt Text on platforms like Mastodon requires a balanced approach. While leveraging community support, direct engagement, and technology each have their benefits, they also present limitations. The future of computer vision holds promise but also necessitates careful consideration of its broader implications. Striking a balance between utilizing AI advancements and maintaining human involvement will be crucial in fostering an accessible and inclusive digital environment.
Comments
Alt text should be built in
Meta, X? And other companies should make it mandatory. Once a pic is posted, before posting is complete, there is no getting around adding alt text to an image, meem etc. However Musk, who broke the hell out of accessibility after purchasing Twitter, plus says he wants nothing to do with it, along with Zuckerberg and other CEOS' don't give a damn whether or not anything is accessible or not. they can blow smoke up your backside all they want about inclusion and everyone's welcome. If you look past that, you'll only get the idea that they keep the culture happy with what they are saying. how many people do you show off to and have them be amazed when you describe a photo you've taken? Or when a celebrity takes the astonishingly kind amount of time to mention you to include something so you don't feel left out? Whether or not AI takes off, it's all down to the specific personality type. some people just don't describe well. Something like, the clouds are blocking the sun. Whereas another person might say, the sun's being shaded by a puffy cotton ball cloud or two and a bird just swooped below them. There's to much focus on this AI junk because right now everyone's trying to be the first. Plus with older devices and the lack of updating for reasons including price, lack of desire to support the company's lackluster stance on anything the individual believes in. I will be keeping AI as far out of any worries as i can. As for alt text, it should be mandatory because the models being used are absolutely deplorable at description.
Can see describe images extention
It maybe in the near future Apple maybe even in iOS 18 with it's onboard AI can seamlessly read any images as a setting you add yourself. It has only been about 4/5 years since suddenly we could interact with Facebook posts images. So I don't see any reason why AI couldn't sit there in the background and via say a sound let you know it is about to describe a photo or image. Could even use a different voice so you know it isn't VO.