ChatGPTを使うにはOpenAIアカウントを作成する必要がある

We’ve trained a model called ChatGPT which interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer followup questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.

https://openai.com/blog/chatgpt

⇧ モデルって言ってるけども、おそらく機械学習のモデルのことを言ってるんでしょうかね？

ChatGPT is a sibling model to InstructGPT, which is trained to follow an instruction in a prompt and provide a detailed response.

https://openai.com/blog/chatgpt

⇧「InstructGPT」の兄弟モデルということで、「InstructGPT」はというと、

openai.com

InstructGPT is better than GPT-3 at following English instructions.

https://openai.com/research/instruction-following

⇧「GPT-3」をより良くしたものということで、「GPT-3」はというと、

Generative Pre-trained Transformer 3 (GPT-3) is an autoregressive language model released in 2020 that uses deep learning to produce human-like text. Given an initial text as prompt, it will produce text that continues the prompt.

https://en.wikipedia.org/wiki/GPT-3

It is the third-generation language prediction model in the GPT series, successor to GPT-2 created by OpenAI, a San Francisco-based artificial intelligence research laboratory.

https://en.wikipedia.org/wiki/GPT-3

GPT-3, which was introduced in May 2020, and was in beta testing as of July 2020, is part of a trend in natural language processing (NLP) systems of pre-trained language representations.

https://en.wikipedia.org/wiki/GPT-3

⇧「機械学習」の中の「深層学習（Deep Learning）」で生成された言語モデルということらしいのですが、

Microsoft announced on September 22, 2020, that it had licensed "exclusive" use of GPT-3; others can still use the public API to receive output, but only Microsoft has access to GPT-3's underlying model.

https://en.wikipedia.org/wiki/GPT-3

⇧「GPT-3」については、Microsoftさんが独占してるってなってますが、「ChatGPT」は『InstructGPT is better than GPT-3 at following English instructions.』って謳ってることから、「GPT-3」より高性能ってことみたいね、あくまで、OpenAIの見解なので、真相は定かではないですが。

Training and capabilities

GPT-3.5

On March 15, 2022, OpenAI made available new versions of GPT-3 and Codex in its API with edit and insert capabilities under the names "text-davinci-003" and "code-davinci-002". These models were described as more capable than previous versions and were trained on data up to June 2021. On November 30, 2022, OpenAI began referring to these models as belonging to the "GPT-3.5" series, and released ChatGPT, which was fine-tuned from a model in the GPT-3.5 series.

https://en.wikipedia.org/wiki/GPT-3

⇧ ChatGPTのリリースは2022年3月15日ということで良いんだろうか？

というか、「ChatGPT」は「GPT-3.5」シリーズのモデルを基盤にしてるってことらしいのだとすると、「GPT-3.5」が「GPT-3」を元にしてる場合は、「ChatGPT」は「GPT-3」を元にしてるということになるってことですかね？

Criticism

GPT-3's builder, OpenAI, was initially founded as a non-profit in 2015. In 2019, OpenAI did not publicly release GPT-3's precursor model, breaking from OpenAI's previous open-source practices, citing concerns that the model would perpetuate fake news. OpenAI eventually released a version of GPT-2 that was 8% of the original model's size. In the same year, OpenAI restructured to be a for-profit company. In 2020, Microsoft announced the company had exclusive licensing of GPT-3 for Microsoft's products and services following a multi-billion dollar investment in OpenAI. The agreement permits OpenAI to offer a public-facing API such that users can send text to GPT-3 to receive the model's output, but only Microsoft will have access to GPT-3's source code.