申請 Google Bard waitlist 見到呢段字,激起我思考到不久嘅將來
當 Offline Generative AI 可以有權限做到 Actionable Task 之時,Rule-based Software 將會很容易被取代
例如寫兩句 Prompt 就做到 2022 年可能要兩三個 Developer 做兩三日嘅功能。
Task 1
- Use this SendGrid API key xxxxxx to send welcome
email to the newly registered users.
- Remember to validate the user email before your send.
- Use the following email templates to draft the email:
"""
You’re on the waitlist
You’ll receive another email when it’s your turn to test Bard.
In the meantime, we asked Bard to write you a little poem while you wait.{{Personalized delightful poem}} ~ Your PA
Thank you for registering.’
"""
Execute Task 1 when there is a new registration completed
event in the message queue.
Connect to the message queue by using the username=xxx, password=xxx, URI=xxxxx.
When executing Task 1, create the email sending history
for audit purpose.
屆時 Generative AI 嘅 Context Length 就係每個獨立 Customized Rule 嘅 Limit。