Generative AI is changing how we write, create images, organize knowledge, and interact with digital systems. But understanding this shift requires more than following new models and product releases. It requires experimentation, critical reflection, and a closer look at the ideas behind the technology.
This blog explores language models, generative art, prompt engineering, associative memory, and emerging AI workflows. It examines how these systems behave in practice, where their limitations become visible, and how they influence creative and intellectual work.
The articles also address the broader questions surrounding artificial intelligence: alignment with human goals, power and responsibility, copyright, regulation, and the increasingly complex relationship between people and machines.
Expect technical explorations, creative experiments, conceptual essays, and practical observations—written for readers who want to understand not only what generative AI can produce, but how it changes the way we think, create, and navigate information.
Seed is a unique attribute of Stable Diffusion-generated images. It represents a specific image and is the master key to it. Same parameters, prompt, and seed always produce the exact same image.
Squared error is a loss function used in machine learning and generative AI to train models to make predictions based on data. MAE reduces average error while MSE does not, and PSNR is no longer considered a reliable indicator of image quality degradation with SSIM emerging as a more suitable metric for assessing image improvements. The squared error loss function is often used in regression tasks and is sensitive to outliers and can be affected by the input/output data scale.
In AI-generated art and storytelling, camera positions can enhance the emotional resonance of your creation. Experimenting with various camera angles can effectively express emotions and support your narrative. This guide explores different camera positions in cinematography, their effects on storytelling and how to adapt them for AI Art and Generative AI platforms.
Estimating the potential outputs of stable diffusion is crucial in understanding its generative capabilities. The size of its latent space determines its capacity for information storage, system complexity, and the probability of copyright collisions. It is challenging to calculate the size of latent space, but having a clear goal in mind when using stable diffusion is essential.
The rise of AI-generated art and NFTs could create a more diverse and accessible art market. However, there are concerns about oversupply and the impact on prices. Despite this, the open nature of AI art and NFTs could reinforce each other and democratize art investment.
AI artists explore the Latent Space, a compressed knowledge model. It refers to all potential outputs allowed by the model. Our creative output, experience, and ideas exist in the infinite latent space.
Hegel was a 19th-century philosopher whose work influenced Karl Marx. NLP experiments with his notoriously complex texts could make them more accessible. Tools like Quillbot and GooseAI can paraphrase and summarize, but ultimately cannot understand the meaning of philosophical concepts.
Transformers are versatile tools for text analysis, including machine translation, summarizing, and paraphrasing. These tasks can be performed through SaaS APIs like OpenAI, finished products like Quillbot, or locally using APIs such as Huggingface Pipeline. By experimenting with text transformation, even a simple sentence can be turned into a short story about the rise and fall of a crypto venture, using temperature parameters and AI tools like GooseAI's playground.
Multiple ML models, like Midjourney, Dalle-E mini, and AI notebooks, enable image and video creation from text input. Google Colab's AI notebooks allow users to tweak the image creation process in detail. However, text-to-image technology has limitations, such as bias and the potential for misuse, which could lead to disturbing or unsettling content.
NLP/NLU transformations lack authenticity and nuances, but Dalle-E images appear perfect. Pre-trained models such as BERT are used to predict outputs, reducing the resources required for AI-based applications. However, pre-trained models have implicit biases and a static world-view, making truth and ethics challenging to achieve.