Microsoft’s Copilot Gets Upgrade to GPT-4 Turbo, Enhancing Language Model and User Experience

Microsoft’s Copilot Upgrades to GPT-4 Turbo: A True Chatbot Upgrade

The new language model brings enhanced capabilities and improved logic

The free version of Microsoft Copilot has taken a significant leap forward with its latest upgrade, transitioning to the more advanced GPT-4 Turbo model, leaving the free version of ChatGPT trailing behind. The improved version of Copilot offers a range of benefits compared to its predecessor, such as a larger language model, footnotes, and access to the internet.

Announced by Mikhail Parakhin, the CEO of Advertising and Web Services at Microsoft, the GPT-4 Turbo has replaced GPT-4 in the free version of Copilot, while Copilot Pro users can still alternate between the two versions. The transition aims to enhance the user experience by fine-tuning the model to ensure at least the same, if not improved, performance compared to GPT-4.

Addressing the limitations

This shift in language model is of paramount importance, as it closes the gap between Copilot and the highly regarded GPT-4 Turbo. GPT-4 Turbo, introduced by OpenAI in November, offers a range of enhancements including an impressive 128,000 context window and an extended knowledge cutoff of December 2023. Previously, however, GPT-4 Turbo was limited to developers via OpenAI’s API or subscribers of Copilot Pro who paid $20 per month. The recent upgrade now brings OpenAI’s advanced model to Copilot’s entire user base free of charge.

Enhanced performance and logical reasoning

To gauge the improvements provided by GPT-4 Turbo, a comparison was made between GPT-4 and GPT-4 Turbo. Inputting the prompt, “I have 5 oranges today, I ate 3 oranges last week. How many oranges do I have left?”, into both models, the results were astounding. GPT-4 failed to understand the logic behind the question, providing the incorrect answer of “2.” Meanwhile, GPT-4 Turbo excelled at reasoning and delivered the correct response. These results indicate that the upgrade to GPT-4 Turbo truly elevates Copilot’s ability to handle logical questions, an essential skill for a chatbot.

Logic and reasoning play critical roles in enabling chatbots to perform a variety of tasks effectively, including mathematics, coding, riddle-solving, and writing, among others. With the advanced capabilities of the GPT-4 Turbo model, Copilot raises its proficiency in these areas, resulting in improved user interactions and a more efficient user experience.