Sources · Used across the app

References

Every claim about real history or research links here. Speculative and conceptual content is labeled as such instead of being sourced — that distinction is the point of this page existing.

1.Turing, A. — “Computing Machinery and Intelligence” (overview)
Turing's 1950 paper posing the imitation game; the conceptual starting point of the timeline.
Open source Used in: Timeline · Symbolic AI
2.Dartmouth workshop (1956) — overview
The 1956 Dartmouth proposal that coined “artificial intelligence” as a field.
Open source Used in: Timeline · Symbolic AI
3.Expert systems — overview (MYCIN, XCON)
Rule-based expert systems of the 1970s–80s: encoded human expertise, brittle outside their rules.
Open source Used in: Timeline · Expert systems
4.Machine learning — overview
Statistical learning from data (1990s–2000s): spam filters, recommendations, hand-engineered features.
Open source Used in: Timeline · Machine learning
5.AlexNet (2012) — overview
Deep convolutional network that won ImageNet 2012 by a large margin; the deep-learning breakthrough moment.
Open source Used in: Timeline · Deep learning
6.Vaswani et al. — “Attention Is All You Need” (2017)
Introduces the Transformer architecture underlying modern large models.
Open source Used in: Timeline · Transformers
7.Brown et al. — “Language Models are Few-Shot Learners” (GPT-3, 2020)
Shows large language models performing tasks from a few examples — the generative-AI inflection.
Open source Used in: Timeline · Generative AI
8.Yao et al. — “ReAct: Synergizing Reasoning and Acting” (2022)
LLMs interleaving reasoning with tool use — a foundation pattern for agentic AI.
Open source Used in: Timeline · Agentic AI
9.Radford et al. — “Learning Transferable Visual Models” (CLIP, 2021)
Joint image–text learning at scale — the technique milestone behind multimodal models.
Open source Used in: Timeline · Multimodal AI
10.Amodei et al. — “Concrete Problems in AI Safety” (2016)
Early research agenda on avoiding side effects, reward hacking, and unsafe exploration — background for the safety section.
Open source Used in: Safety & alignment
11.AI alignment — overview
Survey of the alignment problem: specification gaming, interpretability, and governance directions.
Open source Used in: Safety & alignment
12.Amodei, D. — Anthropic at the UK AI Safety Summit (2023)
Amodei warns advancing capabilities could become “very serious”, including bioweapons assistance; future systems as “country of geniuses in a datacenter”.
Open source Used in: Voices · Frontier leaders
13.Amodei, D. — Anthropic at the Paris AI Summit (2025)
Paris AI Summit statement updating Anthropic's safety/security framing for frontier models.
Open source Used in: Voices · Frontier leaders
14.Fortune — Anthropic CEO on short / medium / long-term AI risks (2023-07-10)
Amodei: short-term bias/misinformation → medium-term capable-model misuse → long-term autonomy/control, “extreme end is concerns about existential risk”.
Open source Used in: Voices · Frontier leaders
15.OpenAI — Introducing Superalignment (2023)
OpenAI (Altman co-author): superintelligence could lead to “disempowerment of humanity or even human extinction”; must be controllable/aligned.
Open source Used in: Voices · Frontier leaders
16.Altman, S. — Testimony before the U.S. Senate (May 2023)
Altman: regulation should incentivize AI safety while preserving benefits; safeguards and oversight for advanced systems.
Open source Used in: Voices · Frontier leaders
17.The Guardian — Musk launches xAI and warns of a “Terminator future” (2023-07-13)
Musk on potentially dangerous advanced AI and xAI's pro-humanity approach.
Open source Used in: Voices · Frontier leaders
18.The National — Musk: AI poses “existential risk” (Nov 2023)
Musk at UK AI Safety Summit: AI an “existential risk”, far more intelligent than us, control uncertain, need a third-party referee.
Open source Used in: Voices · Frontier leaders
19.PBS — DeepMind CEO: “Cautious optimism” (Hassabis interview)
Hassabis: not dismissing risks nor pessimistic about benefits; current systems far from AGI; uncertainty on timelines.
Open source Used in: Voices · Frontier leaders
20.The Guardian — Risk of extinction by AI should be global priority (May 30, 2023)
CAIS joint statement “mitigating extinction risk from AI should be global priority” — signed by Altman, Hassabis, Amodei et al.; Musk not a signatory (separate FLI pause letter). Also carries Hassabis's “take risks as seriously as climate change”.
Open source Used in: Voices · Frontier leaders
21.Microsoft — How do we best govern AI? (May 2023)
Nadella/Microsoft on human-centered AI, governance, safety, transparency, accountability; controlling powerful systems.
Open source Used in: Voices · Frontier leaders
22.Wired — Satya Nadella on AI: “If ... completely out of control, we shouldn't allow it”
Nadella: build safeguards/alignment from the start; do not allow completely uncontrollable systems.
Open source Used in: Voices · Frontier leaders
23.Microsoft — FY2023 Q3 Earnings Call
Context for Nadella's remarks on unintended consequences and built-in safeguards.
Open source Used in: Voices · Frontier leaders
24.Microsoft — Responsible AI Transparency Report 2024
Microsoft's 2024 update on responsible AI governance, safety and transparency practices.
Open source Used in: Voices · Frontier leaders
25.YouTube — AI Worlds: 5 CEOs Agree: AI Could End Humanity? (Sep 15, 2026)
YouTube video (AI Worlds) framing: sequence begins with Jacob Coxon resignation, followed by responses from Amodei, Altman, Musk, Hassabis, Nadella.
Open source Used in: Voices · Sep 2026
26.Anthropic — We Must Pace the Frontier (Sep 2026)
Amodei: pace the frontier; early signs of AI-assisted/recursive self-improvement; OpenAI–Hugging Face incident — stronger safety & independent evaluation.
Open source Used in: Voices · Sep 2026
27.The Guardian — AI safety requires more than just slowing our pace (Sep 15, 2026)
Coverage/analysis of Amodei's pace-the-frontier essay.
Open source Used in: Voices · Sep 2026
28.Ars Technica — AI leaders want to hit the brakes after years of reckless speed (Sep 2026)
Altman agrees pacing warranted; Musk: “Dario is right”; Hassabis: points towards right path + industry standards body.
Open source Used in: Voices · Sep 2026
29.GIGAZINE — AI could take over internet in 6–12 months; pace should slow (Sep 14, 2026)
Nadella: superintelligence must help humanity and remain under human control; support deliberate pacing, benefits broadly distributed.
Open source Used in: Voices · Sep 2026
30.Yahoo Tech — Anthropic researcher resigned because AI could kill us all (Sep 2026)
Jacob Coxon warning: racing toward self-improving superintelligence; AI could potentially kill humanity by end of decade.
Open source Used in: Voices · Sep 2026

← Back to the evolution timeline