Not all waste is revolutionary: a moderate vision of AI

I see a lot of extremely hot AI. It is either completely revolutionary or total garbage. There are not as many people who express a moderate vision of AI. Personally, I think AI is useful in certain situations, but that does not change everything on all my role.

Extreme views

These are not exact quotes, but rather paraphrases that are representative of the declarations that I have read or heard. Have you seen comments like these?

At one end of the spectrum is media threshing on the revolution.

• “The AI ​​will revolutionize everything about how we live and work in the next 2 years!”
• “The AI ​​can create whole courses in a few minutes without any required skill!”
• “The entire software industry will be emptied within 2 years.
• “Anyone in L&D who does not completely change the way he addresses his work will be left behind.”
• “No one will need to create lessons. Everything can be generated by AI on the fly and personalized to the user. ”

At the other end of the spectrum is the idea that everything on AI is garbage, that AI is only a mode that passes.

• “I will never use any AI in my work!”
• “If I see someone sharing an image generated by AI, I block it immediately.”
• “Why are we even talking about AI?” The whole is a grade that will be gone in a year or two like all the other brilliant things. ”
• “Everything that produces is garbage.”
• “I do not know any of the educational designers who really use AI in their work in a substantial way.”
• “Remember when people said we would go up on our smart watches? Ai is the same kind of stupid fashion.”

In the basement of the educational design, a user went so far as to comment on the mods to prohibit all those who mentioned or asked questions about AI. They were so opposed to AI that they didn't even think we should discuss it.

A moderate vision of AI

What I don't see as much is the views in the middle. Of course, there are some – I am several rational people who share more balanced opinions. But these views do not get as much attention as extreme views. Based on my own experience, I have a moderate vision of AI. Although the revolutionary and the garbage exists, I find myself working on the common ground. There are truly practical uses for AI to improve efficiency and increasing capacities at the moment. I can also imagine more important changes on the horizon, and I think we have to look at the long -term possibilities.

Amara's law

“We tend to overestimate the effect of short-term technology and underestimate the long-term effect.”

–Roy Amarafuturistic

I find myself frequently revisiting Amara's law by thinking about AI. Keeping this framework in mind helps me to assess the statements on AI.

Overestimating short -term effects

I see a lot of media threshing on the way in which AI will revolutionize everything in the coming years, making the mistake of the first part of this declaration. Part of the problem is that, whatever the ability of technology that people, organizations and behaviors do not change as quickly.

“Even if the development of AI stopped today, we would have years of change to integrate these systems into our world.”

–Ethan MollickProfessor and researcher in AI

I find myself more agree with Ethan Mollick. The cultural, organizational and societal changes linked to AI will take much more time than what I think most of the predictions indicate. The initial changes with the AI ​​will be small and progressive, regardless of the technology revolution. Organizations move more slowly than the rapid pace of technological change. Technology also continues to improve, and it will take more time for cultural changes to occur.

Underestimating long-term effects

I think the opinions that all of AI is garbage and a fashion that will disappear in a year or two miss the second half of Amara's law. To be clear: a large part of what people produce with AI are garbage. We have probably all seen ridiculous images generated by AI with too many fingers, scary avatars in the Incanny valley and videos generated by AI with broken physics. You have read text that was painfully generated with AI.

But technology improves. IA images are already to the point where they can be indistinguishable from real photos, and AI videos improve quickly. The voices of AI are much better than they were even a year ago, which already considerably affects industry.

It is impossible to really predict the long -term effects of AI. Personally, I am often skeptical about new technological trends. I continue what is happening so I know what is possible, but I generally do not jump as one of the first adopters of technology. Having seen a lot of trends turn off and not be up to the media threw, AI feels different from me. Partly, it is because we can use AI in a practical way at the moment – it is not only something that could be hypothetically useful in the future.

It also looks like a much larger long -term change. In the same way that the Internet has changed many jobs and has created new ones, I think AI could be similar. The Internet is exactly the way we get things done now – communicate, shopping, working. Being online is integrated our lives of hundreds of small ways that differ thirty years ago. In thirty years, AI could be integrated in a similar way into many small ways in our jobs, our education and our communication.

People are also justified to raise ethical and practical concerns with AI. We must think about how we keep a “human in the loop” for most AI uses in order to alleviate potential damage. Energy consumption and environmental effects will require innovation to tackle. I really hope to see more improvements in energy efficiency, not just innovations in energy production. Intellectual property and information security are other major subjects. The AI ​​can allow scams and frauds, some of which are quite frightening. These are all long -term effects that have often been overlooked in AI discussions.

Standing in the middle now

So where does that leave me with my moderate point of view on AI? I am somewhere in the middle. Although I don't think people need to panic if they don't use AI, I think it is intelligent to spend time experimenting to understand how they can help. Even if at the moment, you are just focusing on things where you can improve your efficiency in small quantities, this is probably worth it. Starting with the types of tasks that I shared with my Useful prompts for educational designers is an entry point.

You have to spend time using the tools to have a real idea of ​​what is possible. Ethan Mollick said on several occasions that it takes about 10 hours by working with an LLM like Chatgpt or Claude to be really familiar with the capacities. While the tools improve all the time, you are more likely to find useful LLMs if you work with one of the most recent and better models. If you are on a free LLM and each piece of text looks like the same text that you read everywhere else, the simple fact of changing models can give you better results.

Although I use Claude for many tasks, one of the greatest changes of my own work is in the generation of media. I worked with incredible voice actors, but it becomes more and more difficult to justify the cost of many of my clients. The voices of AI are not as good, but they are good enough for many applications.

For images, I probably use more images generated by AI than stock images these days. It is not only that I can generate the same images as stock images, just faster and cheaper. I can generate unique and personalized images, adapted to specific audiences and objectives. If I need a series of Images in the same style To create a unified look, I don't have to hope to find a good set in an existing library somewhere. I can generate Consistent characters in more natural poses that never possible with images of cut characters. The generation of AI images has limits and problems, but the same goes for stock images.

I am already to the point where I can do more with II images than with stock images, and it is a real improvement in the quality of what I produce. It is not revolutionary, but it is not the garbage either. This is why I find myself with a moderate view.

I also wrote a reflection on How I thought AI will affect educational design In October 2025. Reread this post today, almost a year and a half later, I have the impression that my previous predictions have been reasonably precise so far.

Where are you in the continuum?

Where are you in the continuum of “AI is revolutionary” at “AI is garbage”? Are you somewhere in the middle like me? Do you look more at one end of the spectrum or the other? Let me know in the comments.

Learn more about the way I use AI now

These are some of my messages of “loud work” on how I use AI, focused on practical applications without media threw or fear.

Related