ChatGPT celebrates its first birthday this week, and as a gift, it got more competition! On November 30 exactly one year ago, OpenAI released ChatGPT, around the same time this year Amazon announced Amazon Q and Titan Image Generator.
This is certainly not the first AI competitor to ChatGPT and Dall-E, but for some users it will be super interesting.
Amazon Q is ChatGPT’s answer. For some developers who use AWS, Q will be more interesting than ChatGPT, especially to developers, IT administrators and business analysts. Amazon Q will help write code, answer questions about how to use AWS for administrators, and generate business reports through QuickSight. This can be done on other platforms, but none have the AWS expertise that Amazon has. The company has seventeen years of experience for AWS; Q is built on top of that experience.
Amazon Connect users will be able to use Q directly to help them communicate with customers. Connect is Amazon’s customer support platform. Q will enable analysis of users’ questions and proposal of answers, as well as recommendations for actions or links. After the call ends, Q will generate a call summary that supervisors can review.
This is certainly just a handful of the many possibilities offered by Amazon Q, and we will get more data about their artificial intelligence in the future.
Titan Image Generator
Amazon Q is just one piece of artificial intelligence that Amazon presented at re:Invent. Titan Image Generator stirred quite the interest. Titan Image Generator will enable the creation of realistic, studio-quality images, according to Amazon.
The generation is standard, the user enters phrases of what they want to be generated and gets an image. Titan is not an application but a platform on which developers will be able to develop a solution. The advantage is that Amazon claims to have implemented protections against toxicity and bias, the Verge reported.
In addition to image generation, users will be able to use it for quick and simple background replacement. At the launch, Amazon shared what image generation looks like via the prompt “orange iguana facing right in a rainforest.”
Given that Amazon says their AI generates near-realistic images they had to create some safeguards against using the generated images to share misinformation. They found the solution in adding a “Watermark” to the generated images which will not be limited only to the metadata. So far there is no data on which tools will be able to detect that the image is generated by artificial intelligence.
Amazon will also provide legal protection for users of any artificial intelligence created by Amazon.