Down below you'll find expositions of my random experiments.

A Bit of A Pre-amble on My Thoughts is probably worth reading before going into anything I'm attempting.

GPT3 and Beyond

h/t @alivemachine

Other's Experiments I Find Compelling So Far/Or At Least Have Read:

Writing

GPT-3 Creative Fiction

Questioning/Knowledge

AI2 Leaderboard

GPT3 Encoding Hint I find Useful

In order to just get vectorizations by integers or words...

curl -H "Content-type: application/json" -H "Authorization: Bearer $OPENAI_API_KEY" <https://api.openai.com/v1/engines/davinci/completions> -d '{"prompt":["Hello", "world"]}'
{"id": "cmpl-hEIzploUmImGMr9fZ5N1eLw0", "object": "text_completion", "created": 1595276788, "model": "davinci:2020-05-03", "choices": [{"text": " \\u2013 can you tell me what time the transit centre opens?\\u2019\\n\\n", "index": 0, "logprobs": null, "finish_reason": "length"}, {"text": " code (160 OK), Disable Blacklisting of two-way short code (", "index": 1, "logprobs": null, "finish_reason": "length"}]}

In that case, setting  prompt="Hello world" is equivalent to setting prompt=[15496, 995]

and so I integrated it...