“Over the past decade (2015–2025), AI has rapidly evolved from specialized deep learning models to powerful, general-purpose systems. Early breakthroughs in vision and language (like ResNet and BERT) paved the way for massive language models such as GPT-3 and GPT-4, which brought natural language understanding and generation to the mainstream. Generative AI exploded with tools like ChatGPT, DALL·E, and Stable Diffusion, enabling AI to create text, images, and more. Multimodal models and AI agents began showing reasoning and planning abilities, pushing the boundaries toward AGI. Meanwhile, ethical concerns, regulation, and open-source innovation shaped the global AI landscape.”
The paragraph above was generated by ChatGPT. Within the past decade or so, there has been a rapid increase of development of AI models, and in most recent years, ChatGPT, Dall-E, and many other models designed to generate content or text have been flooding the market. I used to find AI interesting, and there are good use cases for this learning models, but nowadays, AI and it’s creators are seeking to replace humanity in job spaces, in schools, and even in governments. I remember one of the first times I was introduced to AI, it was in a Markiplier video from 2015, where he spoke to the chat bot Eviebot. The point of the video was that is was so hilariously bad and wrong, it was just funny to see what the bot would say or do. Later on, content creators continued to make these same kinds of videos about Talk to Transformer (known more recently as InferKit, which apparently no longer exists? I just looked this up now, it doesn’t exist anymore), and image generators, and the videos were funny because they were bad. It was funny to see fucked up images that vaguely resembled their prompts, and it was sometimes even more funny to see what ai text generators would do based off your prompts. The whole reason the videos were enjoyable wasn’t because they featured good ai, it was because it featured bad ai. Now it has become more refined and capable within the past 2-3 years, and I’ll admit that I do use Chat GPT on occasion (not for school work, with the exception of this post), but it has lost it’s charm that it once had. I can no longer go to these AI models looking for hilarious bad generation, now I get highly detailed pictures and well informed sounding responses. It’s not fun anymore.
Leave a Reply