You might be familiar with the “Technology Camel” model proposed by Elizabeth Ghaffari and Barbara McNurlin, a modification of the traditional Technology Adoption Life Cycle (TALC) which visualizes the adoption of new technologies as a two-humped camel graph.¹


While this image does not explain every wrong attitude towards AI chatbots, it serves as an ever-lasting truth to basically any new technology. There will simply always by a percentage of people who are not willing to dive into the all stuff new tech brings. As the image suggests, over time, they usually adopt it anyway. When ChatGPT released in late 2022, developers had a wild mix of emotions - from excitement to being terrified to death from being replaced by a robot. As any new technology, it was followed by a number of skeptical reviews and opinions, developer all-stars and influencer on LinkedIn surged to provide their invaluable opinion on the matter, as well as trying to calm down the poor folk that thought their jobs became irrelevant. This did not help developers in anyway, rather, it moved them away from the idea of using AI for their daily workflows.

## What using AI means

This is not a praise for AI. I personally don’t have a lot of understanding of how these work, nor do I care - I simply try to evaluate how useful something is to me. With the recent interviews and scandals such as OpenAI deleting their datasets used for training models, it is somewhat concerning, however, this is just my speculation and material for some other time. This article focuses on something I see quite often, and that is the common attitude that using, for example, ChatGPT to figure out a problem or to quickly check how to correctly write a constructor in TypeScript directly means you adhere to shortcuts, you are a bad developer who uses hacks to achieve the said tasks and most importantly, instead of gathering knowledge from books, trying out things yourself thousands of times until it works and similar. While many of the super smart people that built the tools we use, the tech we consume, the programming languages we speak did not have any AI chatbot to help, it does not mean you can’t achieve the same thing while using one. This is essentially the same as if you would like to cook some eggs for breakfast on your stove, but instead why not gather a few logs and rub a stick for hours to get a spark and start a fire, and cook your eggs on a stone instead. Hopefully you understand how absurd it is to reject new, proven technology.

## Using AI properly

Got you there. But I used ChatGPT and Mistral and XYZ and they suck! It can’t solve anything and just repeats stupid things that are not relevant! When I say proven, I mean that it works very well - of course this depends on what exact model you use, how you construct your queries etc, but does it work flawlessly? Hell no. Yeah, it really does suck sometime, especially basic, free models like ChatGPT 3.5. While it works well for most tasks, it often fails to reason what you are trying to ask and quickly shoots at you with long, irrelevant explanations. Over time, things did get better, and will get better - at the time of writing this, Anthropic released Claude Opus 3 - a true state-of-the-art AI that I personally used for about a week. The value you get from using these models is immense. Yes, you might not get the exact answer to your problem, but often times you will get a “search result” far better than you could have ever searched on Google yourself. That’s the whole point - you shouldn’t rely on AI or anything else to solve problems for you, but using AI to boost your work rate and understand issues faster is the key.
For example, instead of pasting the error and saying “fix this for me”, you could ask why did the error occur, explain the background, try to understand the code or the core issue, summarize topics, code review etc. Consider the AI Chatbot as a Senior colleague at your company. 99% of the time, when you ask about an issue you have, they won’t come sit at your desk and do the work for you, rather, they will point you in the right direction by asking key questions or try and provide a possible solution you can try which then leads you in the right direction and so on. You will not become stupid by talking to someone smarter than you every day. However, there is a difference between completely relying on AI to do things for you, versus using it to understand topics, write notes from responses and most importantly - learn.

## Conclusion

Using AI will make you versatile, your knowledge will supercharge, and you will gain insights and directions to further improve on basically anything. Not using AI to solve urgent problems faster, understand complex topics easier and do things better will indeed make you obsolete, you won’t be replaced by AI, but you will be replaced by someone who knows how to use it.