The Race Is On!
Gemini Advanced Challenges ChatGPT-4 for AI Dominance
Google’s new Gemini Advanced model is the first serious challenger to OpenAI’s ChatGPT-4. With our testing, we believe that both are powerful models worth trying out even though there’s still room for improvement with each. Here’s our breakdown…
Both Models:
Cost: $20 a month (free 2 month trial for Gemini)
Have multi-modal input and output
Connected to the internet
Have inconsistent output and hallucinate incorrect responses
Measures have been taken to reduce model bias, but neither have eliminated it
They sometimes exhibit “uncanny behavior” in which it seems like a human is responding
Head-to-Head:
PROs
OpenAI’s ChatGPT-4
Responds faster, especially to complex queries
Has better logical reasoning
Better at complex verbal tasks
Stays on topic & seems to hallucinate less
Ability to upload multiple files in the same chat such as pdfs, png, jpg, webp
Better at data analysis and more complex math tasks
Less restrictions to what requests it will respond to
‘Remembers’ context better in longer chats
Google’s Gemini Advanced
Smoother user experience, fewer errors / glitches
Includes better explanations for its responses
Better at translating languages
Enhanced search capabilities, integrating Google search results, images, & YouTube
Generate near photographic quality images
Can analyze images (location, other details)
More useful integrations (YouTube, Maps, Google Flights, Google Hotels)
Can doublecheck responses through a Google Search option
CONs
OpenAI’s ChatGPT-4
Image generation is less advanced
Less integration with search, which has to be triggered by the user
Less helpful interface
Google’s Gemini Advanced
Image generation is more restrictive, especially of people; not always clear why a prompt is unacceptable
Sometimes forgets its own functionality
Currently does not support file upload other than image
