ClaimsEliezer Yudkowsky

“But it's a giant leap forward based on a technology that was published in, if I recall correctly, 2018.”

Eliezer Yudkowsky, interviewee on Bankless podcast February 2023 Full Transcript: Eliezer Yudkowsky on the Bankless podcast (LessWrong)

Factual AI capabilities

The technology underlying ChatGPT was published, if he recalled correctly, in 2018.

Mostly true True in substance, wrong in a detail.

Factual: Statements about the present or the past, checked against the record: did they hold up?

What would count as true or false

Judged as of 2023-02, allowing for the speaker's hedge ('if I recall correctly'). ChatGPT is built on the transformer architecture and on GPT models. True: the core technology is reasonably identified as published in 2018, for example OpenAI's GPT-1 paper (June 2018) introducing generative pre-trained transformers. Mostly true: the date is defensible but needs a qualifier, for example the transformer itself was published in 2017 while GPT was 2018. Mixed: the underlying technology is just as reasonably dated to 2017 (the transformer) as to 2018, with no clear basis for preferring 2018. Mostly false: the key technology clearly dates to a different year and 2018 matches only a minor component. False: no relevant underlying technology was published in 2018.

Set before the research began.

Evidence (4)

  1. E1 GPT-1 - Wikipedia Wikipedia · 11 August 2021

    “developed following Google's invention of the transformer architecture in 2017.[2] In June 2018, OpenAI released a paper titled "Improving Language Understanding by Generative Pre-Training",[3]”
  2. E2 Generative pre-trained transformer - Wikipedia Wikipedia · 7 February 2023

    “The transformer architecture for deep learning is the core technology of a GPT. Developed by researchers at Google, it was introduced in the paper "Attention Is All You Need", which was released in 2017.”
  3. E3 Generative pre-trained transformer - Wikipedia Wikipedia · 7 February 2023

    “OpenAI was the first to apply generative pre-training to the transformer architecture, introducing the GPT-1 model in 2018.[6]”
  4. E4 Generative pre-trained transformer - Wikipedia Wikipedia · 7 February 2023

    “On June 11, 2018, OpenAI researchers and engineers published a paper called "Improving Language Understanding by Generative Pre-Training", which introduced GPT-1, the first GPT model.[13]”

Reasoning

E3 and E4 show that OpenAI's GPT-1 paper, which introduced the first generative pre-trained transformer, was published in June 2018, so the speaker's hedged 2018 date is defensible. However, E1 and E2 show that the transformer architecture, which E2 calls "the core technology of a GPT," was introduced in 2017. This is the case the criteria list under mostly true: the transformer dates to 2017 and GPT to 2018, so the 2018 date needs that qualifier.

Cite this claim

Eliezer Yudkowsky, “The technology underlying ChatGPT was published, if he recalled correctly, in 2018.” (said February 2023). Verdict: mostly true. Reliability API, claim c_16aa708db15c, method 0.1.0+c876030c, dataset release v1.1.2. https://epistacklabs.com/claims/c_16aa708db15c

This claim is in dataset release v1.1.2, with the same verdict.

Reply to this claim

For Eliezer Yudkowsky, or someone speaking for them: a correction or counter-evidence. Replies are read before anything is shown; an approved reply shows your name, role, date, statement and link. Your email address is never shown and is deleted 30 days after the decision (privacy).

More factual claims by Eliezer Yudkowsky

  1. Resolved true Factual September 2025
    “We learned that that's false, but we did not learn that you can have a thing that can redesign a tree and build an AI much smarter than itself, that builds an AI much smarter than itself, and that…”

    No AI has yet been shown to redesign a tree and build a much smarter AI that builds a still smarter AI; that scenario has not occurred.

  2. Mostly true Factual 25 March 2024
    “I was a passionate noticer of the human brain not being homogenous, and was therefore unpleasantly surprised by transformers scaling as well as they did.”

    Yudkowsky was unpleasantly surprised by how well transformers scaled, having believed the human brain is not homogeneous.

  3. Mostly true Factual 6 April 2023
    “First of all, you realize in principle that the task of minimizing losses on predicting human text does not stop when you’re as smart as a human, right?”

    In principle, the task of minimizing loss on predicting human text does not stop being improvable once a system is as smart as a human.

  4. Mostly true Factual 30 March 2023
    “Transformers are the main thing like that.”

    Transformers were the main architectural change that gave a non-linear jump in performance, compared with earlier approaches such as RNNs.

  5. Mostly true Factual 30 March 2023
    “Because the basic paradigm of machine learning that we are presently operating under is that you can have the loss function, but only for things you can evaluate.”

    The basic paradigm of machine learning currently in use is that a loss function can be defined only for things that can be evaluated.

All of them

Verdicts are made automatically; how verdicts are made.