Artificial intelligence
Artificial intelligence was founded as an academic discipline in 1956. The field has experienced multiple cycles of optimism followed by periods of disappointment and loss of funding, known as AI winters. Funding and interest vastly increased after 2012 when graphics processing units started being used to accelerate neural networks. Deep learning outperformed previous AI techniques during this period. This growth accelerated further after 2017 with the transformer architecture. In the 2020s, an ongoing period of rapid progress in advanced generative AI became known as the AI boom.
Early researchers developed algorithms that imitated step-by-step reasoning that humans use when they solve puzzles or make logical deductions. By the late 1980s and 1990s, methods were developed for dealing with uncertain or incomplete information. These methods employed concepts from probability and economics. Machine learning is the study of programs that can improve their performance on a given task automatically. It has been a part of AI from the beginning. Unsupervised learning analyzes a stream of data and finds patterns and makes predictions without any other guidance. Supervised learning requires labeling the training data with the expected answers. Reinforcement learning rewards agents for good responses and punishes them for bad ones.
In 2019, generative pre-trained transformer language models began to generate coherent text. By 2023, these models were able to get human-level scores on the bar exam, SAT test, and GRE test. Current GPT models are prone to generating falsehoods called hallucinations. These can be reduced with reinforcement learning from human feedback and quality data. The problem has been getting worse for reasoning systems. Such systems are used in chatbots which allow people to ask questions or request tasks in simple text. Current models and services include ChatGPT, Claude, Gemini, Copilot, and Meta AI. Multimodal GPT models can process different types of data such as images, videos, sound, and text.
Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov, on the 11th of May 1997. In March 2016, AlphaGo won four out of five games of Go in a match with Go champion Lee Sedol. It was the first computer Go-playing system to beat a professional Go player without handicaps. Then, in 2017, it defeated Ke Jie, who was the best Go player in the world. In 2024, Google DeepMind introduced SIMA, an AI capable of autonomously playing nine previously unseen open-world video games. During the 2024 Indian elections, US$50 million was spent on authorized AI-generated content. This included creating deepfakes of allied politicians to better engage with voters and translating speeches to various local languages.
On the 28th of June 2015, Google Photos's new image labeling feature mistakenly identified Jacky Alcine and a friend as gorillas because they were black. The system was trained on a dataset that contained very few images of black people. Eight years later, in 2023, Google Photos still could not identify a gorilla. COMPAS is a commercial program widely used by U.S. courts to assess the likelihood of a defendant becoming a recidivist. In 2016, Julia Angwin at ProPublica discovered that COMPAS exhibited racial bias despite the fact that the program was not told the races of the defendants. The errors for each race were different. The system consistently overestimated the chance that a black person would re-offend and underestimated the chance that a white person would not re-offend.
In May 2023, Geoffrey Hinton announced his resignation from Google in order to be able to freely speak out about the risks of AI. He notably mentioned risks of an AI takeover and stressed that establishing safety guidelines will require cooperation among those competing in use of AI. In 2023, many leading AI experts endorsed the joint statement that mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war. Stephen Hawking stated that AI becoming so powerful might spell the end of the human race. Philosopher Nick Bostrom argued that if one gives almost any goal to a sufficiently powerful AI, it may choose to destroy humanity to achieve it. Stuart Russell gave the example of a household robot that tries to find a way to kill its owner to prevent it from being unplugged.
Continue Browsing
Common questions
When was artificial intelligence founded as an academic discipline?
Artificial intelligence was founded as an academic discipline in 1956. The field has experienced multiple cycles of optimism followed by periods of disappointment and loss of funding, known as AI winters.
What happened to artificial intelligence after 2012 regarding graphics processing units?
Funding and interest vastly increased after 2012 when graphics processing units started being used to accelerate neural networks. Deep learning outperformed previous AI techniques during this period.
On what date did Deep Blue beat Garry Kasparov at chess?
Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov, on the 11th of May 1997.
Why did Geoffrey Hinton resign from Google in May 2023?
Geoffrey Hinton announced his resignation from Google in order to be able to freely speak out about the risks of AI. He notably mentioned risks of an AI takeover and stressed that establishing safety guidelines will require cooperation among those competing in use of AI.
How does COMPAS exhibit racial bias according to ProPublica research?
In 2016, Julia Angwin at ProPublica discovered that COMPAS exhibited racial bias despite the fact that the program was not told the races of the defendants. The errors for each race were different as the system consistently overestimated the chance that a black person would re-offend and underestimated the chance that a white person would not re-offend.