AI shipped faster, met its first binding law, and won two Nobel Prizes

Frontier AI settled into a release cadence in 2024. OpenAI, Google, Anthropic, and Meta each shipped multiple model families, Anthropic’s Claude 3 Opus broke OpenAI’s run at the front, and in September OpenAI’s o1 made reasoning before answering a second axis of competition alongside sheer scale. Governments moved from statements to law: the European Union’s AI Act entered into force in August, sixteen companies signed voluntary safety pledges in Seoul, and California’s governor vetoed the furthest-advanced American attempt to regulate frontier developers. A cloned presidential voice reached New Hampshire voters by telephone in January. In October, Nobel committees honored neural-network research in physics and AlphaFold in chemistry.

23 entries recorded

November

  1. Amazon doubles its Anthropic investment to $8 billion

    Amazon said on 22 November 2024 that it would invest a further $4 billion in Anthropic, taking its total commitment to $8 billion while remaining a minority investor. The deal named AWS Trainium as Anthropic’s main training chip and expanded joint work on that chip.

October

  1. Family sues Character.AI after a teenager’s death

    Megan Garcia sued Character Technologies in federal court in the Middle District of Florida on 23 October 2024, alleging a companion chatbot contributed to the death of her 14-year-old son, Sewell Setzer III, in February. Google and Alphabet were also named as defendants.

  2. Nobel chemistry prize honors AlphaFold and David Baker

    The Royal Swedish Academy of Sciences gave half of the 2024 Nobel Prize in Chemistry to Demis Hassabis and John Jumper of Google DeepMind for protein structure prediction with AlphaFold, and half to David Baker of the University of Washington for computational protein design.

  3. Nobel physics prize goes to Hopfield and Hinton

    The Royal Swedish Academy of Sciences awarded the 2024 Nobel Prize in Physics to John Hopfield and Geoffrey Hinton on 8 October 2024 for foundational discoveries that enable machine learning with artificial neural networks. They shared 11 million Swedish kronor.

September

  1. Newsom vetoes California’s SB 1047 frontier AI bill

    Governor Gavin Newsom vetoed SB 1047 on 29 September 2024, a bill requiring developers of large AI models to run safety tests and build in a shutdown capability, which had passed the Assembly 48-16 and the Senate 30-9 in August. He asked experts to help draft guardrails instead.

  2. Mira Murati leaves OpenAI

    OpenAI’s chief technology officer Mira Murati said on 25 September 2024 that she was leaving after six and a half years, wanting “the time and space to do my own exploration.” Chief research officer Bob McGrew and research vice president Barret Zoph left the same week.

  3. OpenAI releases o1, its first reasoning model

    OpenAI released o1-preview and o1-mini on 12 September 2024, models trained to work through a hidden chain of reasoning before answering. OpenAI reported that o1 scored 83% on a qualifying exam for the International Mathematical Olympiad, against 13% for GPT-4o.

August

  1. The EU AI Act enters into force

    The European Union’s AI Act entered into force on 1 August 2024, twenty days after publication in the Official Journal. The law sorts AI systems into four risk tiers, bans some uses outright, and phases in over two years, with most high-risk rules applying from August 2026.

July

  1. DeepMind systems reach silver-medal standard at the IMO

    Google DeepMind said its AlphaProof and AlphaGeometry 2 systems solved four of six problems at the 2024 International Mathematical Olympiad, scoring 28 of 42 points, one short of a gold medal. The systems took up to three days per problem against contestants’ 4.5 hours.

  2. Meta releases Llama 3.1 with a 405-billion-parameter model

    Meta released Llama 3.1 on 23 July 2024, including a 405-billion-parameter model it called the largest openly available foundation model to date, with a 128,000-token context window. Meta said it matched GPT-4o and Claude 3.5 Sonnet on knowledge, math, and tool-use benchmarks.

June

  1. Record labels sue Suno and Udio over training data

    Sony Music Entertainment, UMG Recordings and Warner Records, coordinated by the RIAA, sued Suno in Massachusetts federal court and Udio’s Uncharted Labs in the Southern District of New York on 24 June 2024, alleging both trained on copyrighted recordings without a license.

  2. Ilya Sutskever founds Safe Superintelligence

    Ilya Sutskever announced Safe Superintelligence Inc. on 19 June 2024, a month after leaving OpenAI, with Daniel Gross and Daniel Levy. The founding statement named “one goal and one product: a safe superintelligence” and gave offices in Palo Alto and Tel Aviv.

  3. Nvidia becomes the world’s most valuable public company

    Nvidia’s market capitalization reached $3.33 trillion on 18 June 2024, passing Microsoft’s $3.31 trillion to make it the most valuable listed company in the world. Its shares rose 3.7% to $135.77, after a ten-for-one split on 7 June.

  4. Apple announces Apple Intelligence at its developer conference

    Apple introduced Apple Intelligence at its Worldwide Developers Conference on 10 June 2024, a system built into iOS 18, iPadOS 18, and macOS Sequoia that writes text, makes images, summarizes notifications, and can pass complex questions to ChatGPT with the user’s permission.

May

  1. Sixteen companies sign safety pledges at the AI Seoul Summit

    Sixteen companies, among them OpenAI, Google, Microsoft, Anthropic, Meta, and Mistral AI, signed the Frontier AI Safety Commitments at the AI Seoul Summit on 21 May 2024, agreeing to publish safety frameworks and to hold back models whose severe risks they could not mitigate.

  2. Scarlett Johansson says OpenAI’s Sky voice copied hers

    Scarlett Johansson said on 20 May 2024 that she had twice declined OpenAI’s requests to voice ChatGPT and was “shocked, angered and in disbelief” at Sky, a voice listeners compared with hers. OpenAI paused Sky and said it had cast its actor before contacting her.

  3. OpenAI launches GPT-4o, a single model for text, audio, and images

    OpenAI introduced GPT-4o on 13 May 2024, one model trained across text, image, and audio, demonstrated by chief technology officer Mira Murati in a livestream showing real-time voice conversation. OpenAI gave it to free ChatGPT users and said it handled around 50 languages.

  4. DeepMind and Isomorphic Labs unveil AlphaFold 3

    Google DeepMind and Isomorphic Labs announced AlphaFold 3 on 8 May 2024, extending the protein-structure predictor to complexes of proteins with DNA, RNA, small-molecule ligands, and ions. The Nature paper reported at least a 50% accuracy gain over prior methods.

April

  1. Meta releases Llama 3 under an open license

    Meta released the first Llama 3 models, at 8 billion and 70 billion parameters, on 18 April 2024 under a license permitting most commercial use. Meta said the 70B model beat comparable models, including Claude 3 Sonnet, on human evaluations across twelve use cases.

March

  1. Anthropic launches the Claude 3 model family

    Anthropic released Claude 3 Opus, Sonnet, and Haiku on 4 March 2024, three models spanning a range of speed and capability, with a 200,000-token context window and image input. Anthropic reported that Opus matched or exceeded GPT-4 on several public benchmarks.

February

  1. Google pauses Gemini image generation of people

    Google stopped Gemini from generating images of people on 23 February 2024 after it produced racially diverse depictions of Nazi-era German soldiers and the US founding fathers. Senior vice president Prabhakar Raghavan said tuning for diversity had overgeneralized.

  2. OpenAI previews Sora, a text-to-video model

    OpenAI previewed Sora on 15 February 2024, a model that generates up to a minute of high-definition video from a text prompt. Access went first to red-teamers and a group of visual artists; a standalone product followed in December 2024.

January

  1. AI-cloned Biden robocall targets New Hampshire primary voters

    A robocall using an AI-generated imitation of President Joe Biden’s voice told New Hampshire voters not to vote, two days before the state’s 23 January 2024 presidential primary. Political consultant Steve Kramer admitted commissioning it; the FCC later fined him $6 million.