OpenAI Enhances ChatGPT to Blur Writer-AI Distinction

OpenAI has introduced an improved version of GPT-4 Turbo, a model that powers its conversational AI tool ChatGPT. The upgraded version aims to enhance the tool’s ability to engage in more natural conversations. This latest version, named “gpt-4-turbo-2024-04-09,” offers enhancements in writing, math, logical reasoning, coding, and has an updated knowledge base. It has been trained on publicly available data up until December 2023, compared to the previous version, which only had access to information up until April 2023. Premium users subscribed to ChatGPT Plus, Team, or Enterprise plans can now utilize this updated version.

OpenAI addressed the issue of verbose and less conversational responses in their post on X, stating that the new GPT-4 Turbo will result in more direct and conversational language. This update also brings concerns about the challenge of distinguishing between human and AI-generated content. Paul Graham, co-founder of Y Combinator, ignited a discussion when he claimed that an email containing the word “delve” must have been composed with the assistance of AI. This sparked a conversation about different regions using English vocabulary in diverse ways, implying that writers who speak English as a second language may face challenges as their work may resemble AI-generated content.

In addition to the GPT-4 Turbo upgrade, OpenAI recently launched new models in its API, including GPT-4 Turbo with Vision, which incorporates image understanding capabilities. To combat what developers refer to as “laziness” in the model’s task completion, OpenAI quietly updated the training dataset for GPT-4 Turbo to include relevant information up until December 2023. OpenAI eliminated the requirement for users to create an account to access ChatGPT3.5, their widely used generative AI tool. This change aims to make it easier for users to experience the potential of AI. Users without accounts will not be able to save their previous interactions with the tool.