Yudkowsky said alarm bells went off for him in 2015, when it became obvious to him "this is how it was going to go down"; he said this in answer to a question about the large amount of capital going into AI investment.

This is a claim about Yudkowsky's own state of mind: in 2015 he became alarmed that AI development would be driven by heavy capital investment and large-scale industry efforts. True: his writings or statements from 2015 to early 2016 show that heightened alarm, or a stated change in his view, about this path (for example, reactions to the founding of OpenAI or DeepMind's scaling, or to capital flowing into deep learning). Mostly true: such alarm is documented, but the turning point is placed in late 2014 or 2016, or the concern is framed somewhat differently, for example as being about deep learning progress in general. Mixed: the record shows some concern in 2015, but it is unclear whether this was a turning point. Mostly false: his contemporaneous writings place his alarm at a clearly different time or for clearly different reasons. False: his 2015-era writings show him not alarmed, or expecting AI development to go some other way.

Set before the research began.

1. E1 Yudkowsky on AGI risk on the Bankless podcast - Machine Intelligence Research Institute Machine Intelligence Research Institute · 14 March 2023 Own record

   “That was like… that was me in 2015 going like, “Oh. So this is what humanity will elect to do. We will not rise above. We will not have more grace, not even here at the very end.””

2. E2 When did Eliezer Yudkowsky change his mind about neural networks? — LessWrong LessWrong · 14 November 2023 Own record

   “Sometime between 2014 and 2017, I remember reading a discussion in a Facebook group where Yudkowsky expressed skepticism toward neural networks.”

3. E3 When did Eliezer Yudkowsky change his mind about neural networks? — LessWrong LessWrong · 14 November 2023 Own record

   “My guess is AlphaGo-- I once heard someone who worked at MIRI say that they watched the event and Eliezer was surprised by it.”

4. E4 Yudkowsky on AGI risk on the Bankless podcast - Machine Intelligence Research Institute Machine Intelligence Research Institute · 14 March 2023 Own record

   “That was like… that was me in 2015 going like, “Oh. So this is what humanity will elect to do. We will not rise above. We will not have more grace, not even here at the very end.””

5. E5 MIRI’s 2025 Fundraiser — EA Forum Effective Altruism Forum

   “About the unexplained shift of focus from symbolic AI, which Yudkowsky was still claiming as of around 2015 or 2016 — quite late in the game, all things considered — was more likely than deep learning to lead to AGI, to deep learning:”

6. E6 MIRI’s 2025 Fundraiser — EA Forum Effective Altruism Forum

   “This is made up, as far as I can tell (at least re: symbolic AI as described in the wikipedia article you link).”

7. E7 OpenAI and other news - Machine Intelligence Research Institute Machine Intelligence Research Institute · 11 December 2015 Own record

   “I’m excited to see OpenAI joining the space, and I’m optimistic that their entrance will result in promising new AI alignment research in addition to AI capabilities research.”

8. E8 OpenAI and other news - Machine Intelligence Research Institute Machine Intelligence Research Institute · 11 December 2015 Own record

   “My feelings about OpenAI at the time were, IIRC, some cautious optimism plus a bunch of pessimism.”

9. E9 Podcast/video/transcript: Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality EA Forum · 9 April 2023 Own record

   “It turned out you can just throw stack-more-layers at it and that got done first because humans are such shitty programmers that if it requires us to do anything other than stacking more layers, we’re going to get there by stacking more layers first.”

The only direct support is Yudkowsky looking back in 2023 (E2, E6), where he describes a 2015 reaction of dismay at "what humanity will elect to do." That is the subject restating his own claim, not a record from the time. The December 2015 MIRI post on OpenAI (E9) expresses optimism, and a statement on the same page (E10) recalls "cautious optimism plus a bunch of pessimism" at the time, but the quotes do not show that these are Yudkowsky's own 2015 words about heavy capital investment. The forum and LessWrong items (E3, E4, E7, E8) are unsourced recollections or disputed assertions about his views on neural networks, and none of them sets a date for his alarm.