Chat replay
hello everyone it's happy Wednesday it was very long day for me I had a four meetings in particular for the second meeting I have the opportunity to meet one of the partners for me immersive Unity with discussed how we can collaborate together Thunderbird XR platform and services working together with immersive Unity to host a summer hiker song 2023 in summer between June and July platforms using Sandra Burden XR platform in the services as the Baseline Okay so I was being receiving quite a few is people talking about hey you guys great you're integrating with check GPT already check GPT going to be the expert for everything it's a great to see Thunderbird can integrated with the check GPT and also you guys will get prepared for barn Google Barn integration the ga release next month I'll see this way I might really upset some friends and whoever love should you think you can do everything a lot of different work different industry and business sectors can you can do that with chat GPT but to upset you with this very quick and a short demo I'm going to show you there is a fundamental differences between check GPT and GPT model and how the GPT transcriber Transformer being trained what they could do and what they could not do well on the other side Hope barnai Has would be able to cover what Jeff GPT not able to do by Design right okay so everybody read a lot of articles about the GPT model right if you read it how the GPT based Transformer being created trained you should be able to understand reinforced training and learning again right so based on that diagram you should know GPT is not feed model for any kind of conversation or dialogue is that need real time data feeds it's not real time right it just cannot absorb the real-time data feed January it's a real-time day to feed base the response while on the other side Google Barney is a Lambda is there Lambda model based Lambda model by Nature it is a building real-time data feeds so besides supervised learning optimize reinforcement learning I believe they took reinforcement learning and transfer learning too which is applicable for real-time therapy based transcriber or Transformer what do you want to call so to put a very simple example right what I did is so I'm not going to show how our app works we wish our app is basically it took my speech transcribe to tax then we do the alternate LP analyze the text then generate the command the command is going to fit into our Beauty our bot the bot going to response back which is real time responses right so what I just did today just now right I going to give us some up thanks for being Frank all right what I did is just I I type it in or use my voice really doesn't matter I was looking for Australian dollar price in Swiss franc so in our app will trans which one's crab this speech to text and analyze this text generate a command which is which will be this Command right Judy CHF eventually it's been triggered in our cloud so this kind of real-time feed quote a UD with the base currency is Frank X3 can be blah blah blah blah this look at what a check GPT response back seat hey I'm sorry but I say AI language model I do not have access to real-time Market data however you can use it to check this current exchange rate between the Australian dollar and Swiss franc using a reliable Financial website or app such as XE Yahoo finance Google Finance alternatively you can contact your bank and financial institutions for most up-to-date information on the currency exchange rate okay there are two things I want to point it out right now how to ultimately transcribe bring to the second Point meaning check GPT and all GPT 3.x model it's a snapshot meaning they are knowledge base their ability they are wonderful responses are all based on Snapchat data feet that all all the current GPT Transformers being trained for all right so I hope this this video answer love your concerns and questions but again I wish you a wonderful Wednesday we'll see you in the next time