Best Guinness World Records in AI


Artificial intelligence is developing at record speed, literally. Through exponential development, AI has made its way into the Guinness Book of World Records. Below is a list of records in the AI ​​domain.

The greatest lesson in AI programming

What started as a simple BotCamp program has become a world record for the greatest artificial intelligence programming lesson. Capital One Services LLC hosted this camp as part of its Future Edge DFW initiative in Dallas, Texas, USA on April 17, 2019.

The AI ​​part of the lesson was part of a longer coding lesson that lasted over 90 minutes. The lesson ended with the construction of a Dallas Cowboys celebrity bot Jason Witt (who was present). The lesson had a pass rate of 94.8%.

First video game with AI

Metal Gear Solid 2, the Japanese video game developed by Konami Computer Entertainment, was the first stealth game with AI features and NPCs that could form teams, communicate with each other, and counterattack. Disabling these NPCs before reinforcements arrived was an important part of gameplay, and this level of AI has since become part of the stealth subgenre. GWR Video Gamer’s Edition 2008, title: First collective AI in a stealth game. The game won the world record as the first stealth game to use AI on such a scale.

Most people complete an online AI lesson within 24 hours

On October 14, 2020, Intel India in collaboration with CBSE created a Guinness World Record by hosting a online course on artificial intelligence entitled “demystifying the impact of AI” which received 12,701 participants from grades 8 and above in 24 hours. The virtual classroom was part of their “AI For Youth Virtual Symposium”.

First gender-aware robot

On January 20, 2003, Intelligent Earth developed a robot Doki with visual gender recognition software. Doki uses high-resolution cameras and computer vision software to detect head movements, etc. and identify the gender of females with 100% accuracy and males with 96% accuracy.

First political chatbot

Entrepreneur Nick Gerritsen worked with Touchtech and Victoria University of Wellington to develop the Semantic Analysis Machine (SAM). SAM holds the world record as first political chatbot and is designed to answer questions about specific issues through Facebook Messenger.

The chatbot can converse on various topics ranging from housing to education. In addition to helping people understand various policies, SAM also raises awareness about artificial intelligence and how public opinion is influenced by social media.

First AI Scientist

Eureqathe proprietary modeling engine created by Cornell’s Artificial Intelligence Lab (Creative Machines lab), holds the world record for the first AI-based scientist. Invented in 2009, this program is able to analyze data on any subject and generate a mathematical law, assimilating the given data. Named after the famous expression of Archimedes, this program works on the principle of the genetic algorithm – replicating the perfecting power of evolution in the natural world.

As a proof of concept, the team fed the program data on the motion of a pendulum – which they responded to by rediscovering Newton’s second law of motion and the conservation of energy.

First computer film critic

UK based Epagogixfounded in 2003, uses neural networks and analytics software to predict a movie’s box office success.

The tool takes data from human scores on the prospective script, compares it with data from previous film scripts, and predicts what a full production of the new script is likely to produce. As a result, the tool can make an accurate estimate of a movie’s box office with a margin of +/- 10 million USD.

The biggest stock market crash caused by automated trading

The $1 trillion flash crash of May 6, 2010 lasted about 36 minutes. As a result, the US stock market fell more than 600 points, only to rally again within 20 minutes.

Algorithmic trading (AI computers automatically executing trades by pre-programmed codes) caused the crash, and it holds the record for the biggest stock market crash triggered by AI.

The most expensive AI-generated artwork sold at auction

Edmond de Belamy, the 2018 Generative Adversary Network-based portrait painting by Parisian arts collective Obvious, holds the registration for the most expensive work of art sold at auction created by a machine. The job brought in $432,000 (£334,144). The algorithm was based on the work of Robbie Barrat, an open source AI programmer.

Best score in a live Turing test

The highest score in a live (as opposed to online) and controlled Turing test was achieved by a chatterbot called Cleverbot, written by British computer scientist Rollo Carpenter. The test took place at the Techniche festival in Guwahati, India on September 3, 2011. Thirty conversations with clever robot were observed and rated for their human appearance on a scale of 0 to 10 – anything above 5 (or 50%) is a pass. Out of a total of 1,334 votes cast by the judges, clever robot achieved a score of 59.3% human.

Best score in the CASP contest

In November 2020, AplhaFold 2 achieved the highest Critical Assessment of Protein Structure Prediction (CASP) score. The AI-based system created by DeepMind scored 92.4 GDTs at CASP14 (which ran from May to September 2020).

In 2020, Alpha folding scored a GDT of 87.0 in the most difficult category of “free modelling” comprising the most difficult and complex protein problems. The AlphaFold system uses a bank of 16 TPUs (Tensor Processing Units) and responds within weeks. The set of 16 TPUs is roughly equivalent to 100-200 commercially available GPUs.

Most Loebner Awards win

Stephen Worswick has won the most Loebner Awards. His chatbot Mitsuku won the Loebner Prize in 2013, 2016, 2017, 2018, and 2019. The Loebner Prize is awarded to a reputable most humane artificial intelligence computer program.

First inventor of AI

DABUS (Device for the Autonomous Bootstrapping of Unified Sentience]) is an artificial intelligence (AI) system created by Stephen Thaler.

In July 2021, DABUS and Thaler were listed as inventors on South African patent ZA2021/03242 “Food Container and Devices and Methods for Attracting Enhanced Attention”.

DABUS works using a pair of adversarial neural networks called the inventor and the critic. The former comes up with designs, while the latter evaluates those designs and changes settings to fine-tune the former’s output.

The two inventions he proposed for this patent were a plastic food packaging design that used fractal geometry to improve heat transfer within the packaging and a pulsed-light “neural flame” designed to rapidly attract heat. attention of people in noisy and busy environments.


Comments are closed.