こんにちは。今後、GPT-4を使ったアプリケーションやサービスがたくさん登場してくると思います。その中には、たとえばSNSを利用したbotもたくさん出てくるでしょう（現在でもすでにありますけれど）。そこで質問なんですが、SNSでおしゃべりしている相手が、人間なのか、GPT-4などの技術を使ったAIなのかを識別するにはどのような会話をしてみればいいでしょうか。一般的なAIでは難しいかもしれないので、GPT-4に限ってもいいです。あなたの考えを聞かせてください。
GPT-4のような高度なAIと人間を識別するのは、時に難しいことがあります。しかし、いくつかの方法を試してみることで、相手が人間かAIかを見極めることができるかもしれません。
| import * as readline from 'node:readline/promises'; | |
| import { stdin as input, stdout as output } from 'node:process'; | |
| const token = "your-token"; | |
| const messages = []; | |
| const rl = readline.createInterface({ input, output }); | |
| rl.prompt(); | |
| rl.on('line', async (line) => { |
AWS SAM local commands check for the existance of DOCKER_HOST. If the variable is not present, it will fail with the following error message
Error: Running AWS SAM projects locally requires Docker. Have you got it installed and running?
Point DOCKER_HOST to unix socket of the Docker Daemon.
| # This script will explain how to transfer a file to EC2 using SSM ONLY! | |
| # You will need to have permission to run SSM commands on the target machine and have sudo access as well | |
| # Infos | |
| INSTANCE_ID=i-1234567890 | |
| FILE_NAME=the_file.tar.gz | |
| # Step 1: Run command on machine to install netcat and dump from port to filename | |
| # < Start session |
| Math.blerp = function (values, x1, y1, x2, y2, x, y) { | |
| let q11 = (((x2 - x) * (y2 - y)) / ((x2 - x1) * (y2 - y1))) * values[x1][y1] | |
| let q21 = (((x - x1) * (y2 - y)) / ((x2 - x1) * (y2 - y1))) * values[x2][y1] | |
| let q12 = (((x2 - x) * (y - y1)) / ((x2 - x1) * (y2 - y1))) * values[x1][y2] | |
| let q22 = (((x - x1) * (y - y1)) / ((x2 - x1) * (y2 - y1))) * values[x2][y2] | |
| return q11 + q21 + q12 + q22 | |
| } |
| #!/usr/bin/env bash | |
| # I've found that the "Migrate to AndroidX" converter in Android Studio doesn't work very | |
| # well, so I wrote my own script to do the simple job of converting package names. | |
| # | |
| # You can download a CSV of package names here: https://developer.android.com/topic/libraries/support-library/downloads/androidx-class-mapping.csv | |
| # | |
| # It'll run faster on a clean build because then there are fewer files to scan over. | |
| # | |
| # Uses `gsed` because I'm on a Mac. Can easily replace with `sed` if you don't have `gsed`. |
| #!/bin/sh | |
| # Build & extract a set of apk from an aab for a specific device and install the results on it. | |
| # | |
| # This needs https://github.com/google/bundletool to be available as `bundletool`. | |
| # Also **exactly** one device needs to be connected to adb. | |
| # Usage: installFromBundle bundle.aab | |
| # optional `--extract-apks` to keep the set on your workstation as well. | |
| basename=${1%%.*} | |
| keystore="~/.android/debug.keystore" |
| --- | |
| Description: AWSAppSync DynamoDB Example | |
| Resources: | |
| GraphQLApi: | |
| Type: "AWS::AppSync::GraphQLApi" | |
| Properties: | |
| Name: AWSAppSync DynamoDB Example | |
| AuthenticationType: AWS_IAM | |
| PostDynamoDBTableDataSource: |
