Natural language processing and machine learning are two of the artificial intelligence disciplines that are seeing greater development. At their most rudimentary level, NLP constructs interpret data input by humans and pass them onto ML constructs to deliver results; one example would be Google search queries that are processed semantically to deliver information sought by humans. Could these AI systems be used for web design?
Sharif Shameem is the founder of Debuild, an interesting tech startup that aims to let non-coders build apps by means of describing them. Shameem has been playing with GPT-3, a language model that is part of the OpenAI consortium, and which can handle about 175 billion parameters. GPT-3 is all about NLP, and to get an idea of just how advanced and powerful it has become, it helps to know that GPT-2 could handle 1.5 billion parameters. The Microsoft Transformer project, which was introduced with great fanfare earlier this year, could handle 17 billion parameters.
When Microsoft released its Transformer project a couple of months ago, journalists and content authors were rightfully concerned, and it is now the turn of web designers to feel this way. For more information click here https://twitter.com/sharifshameem/status/1282676454690451457.