This episode is brought to you by the Cloud Wars Expo. This in-person event will be held June 28th to 30th at the Moscone Center in San Francisco, California.
Highlights
00:15 — GPT is a neural network that’s used for natural language processing (NLP). With GPT3, we can use the network and computers to produce human text or speech.
00:48 — In the Reddit community, there are two bots having a conversation with each other by using GPT3 to create the text.
01:12 — The success of GPT3 can be, in part, credited to the expansion in the number of layers and parameters that can be used in the network. So, they can pack the neural network with all the possibilities required to produce natural human text.
01:35 — The main difference with GPT3 is that there are simply more parameters—there’s no new NLP algorithm and no new neural network technology.
02:08 — This technology is good for short communication, as it tends to fall apart after about a paragraph of text.
02:41 — The ability to create, read, and comprehend human text and then develop an output response to human text is incredibly powerful. This can lead to the creation of even more solutions.
03:10 — The GPT3 technology is owned and controlled by OpenAI. Microsoft is heavily invested in OpenAI. One of the use cases they have found is that it works really well in auto code.
04:20 — However, we still need to understand that there are limitations to this technology.
Looking for real-world insights into artificial intelligence and hyperautomation? Subscribe to the AI and Hyperautomation channel:
