-
Artificial Intelligence (AI)
Artificial Intelligence (AI) has been a sought-after goal since the early 20th century, as depicted in movies as early as the Tin Man in The Wizard of Oz. But how has that early goal been realized in today's reality? -
The Turing Test
In 1950, mathematician Alan Turing published his paper "Computing Machinery and Intelligence." The paper asked whether or not computers could think, and then developed a test to determine whether or not they could. To pass the test, a computer would need to imitate conversational written human language that is not distinguishable from that of a human's. While Turing presents contradictions to this test's premise in his paper, a computer that could pass would be considered able to think. -
Logic Theorist
About 5 years after Turing published his paper, Allen Newell and Herbert Simon created the Logic Theorist computer program. This program simulated human problem-solving skills. The program was heavily funded and was also presented at the Dartmouth Summer Research Project on Artificial Intelligence. -
Dartmouth Summer Research Project on Artificial Intelligence
John McCarthy and Marvin Minsky hosted the Dartmouth conference in hopes of dynamic discussions of AI. Although the conference did not go as planned, it still acted as a catalyst for the field of AI, and coined the term "Artificial Intelligence" and "AI" for the first time. -
Expert Systems
Edward Feigenbaum became the father of expert systems due to his research at Stanford University from 1965 to 1982. Expert systems are AI programs that mimic decision-making for a very specific domain of information, such as medical diagnosis, or chemistry properties. The first expert system was created by Feigenbaum and his colleagues. It was called the DENDRAL project. -
SYSTRAN Translation Services
Many translation tools began their development in the late 1960s and early 1970s. Beginning as a dictionary-based tool created by Peter Toma in 1968, SYSTRAN, which stands for "System Translation," has become a premier translating service. Watch a video produced for their 50th anniversary here. -
NETtalk
NETtalk is an AI program that was created to map the patterns of English text input and phonetic transcriptions to be a text-to-speech program. The program could not visually detect text on its own, but it did demonstrate AI pattern recognition and ability. -
Chess Champion
By the end of the 1990's AI is sophisticated enough to beat chess masters at their own game. World chess champion Gary Kasparov was beat by IBM's Deep Blue chess-playing program. This was a large accomplishment towards showing the decision-making capabilities of AI computer systems. Watch a short video about Deep Blue's win here. -
Speech Recognition Software
In the same year that Deep Blue beat the chess grand master, Windows implemented a speech recognition program in their software. Created by Dragon Naturally Speaking, the program was an early speech-to-text tool. -
Social Robots
For a machine to truly be able to think like a human, it should be able to understand human emotion. Cynthia Breazeal created the robot Kismet to be socially intelligent. The robot responds to stimuli and can communicate various emotions. The algorithm also allows the robot to learn from its mistakes. Watch a video featuring Breazeal and Kismet here. -
Self-Driven
The Defense Advanced Research Projects Agency (DARPA) issued a challenge in the early 21st century to create a self-driving vehicle that could travel unknown terrain for several miles. During the first challenge in 2004, no robots could complete the course. In the repeated challenge in 2005, 23 vehicles raced, but only 5 finished the course. Stanford's robot "Stanley" finished first and won the $2 million prize. -
IBM is Back to Play Again!
Beating the world champion at chess wan't enough for IBM. In 2011, IBM's Watson competed and won Jeopardy! while beating two previous champions. Watson's program is structured to find answers in unstructured data significantly quicker than standard search engines. Watch Watson compete here. -
Siri
As a speech-recognition softwares were being developed throughout the late twentieth and early twenty-first century, the programs were not user-friendly enough to be used on a daily basis. Apple turned that fact on its head with the introduction of Siri with the iPhone 4S. Siri uses speech and AI technology to assist iPhone users in helpful ways. Siri was leading in personal AI technology. -
AlexNet: Image Recognition
AlexNet, designed by Alex Krizhevsky, is an image-recognition Convolutional Neural Network (CNN). AlexNet won the 2012 ImageNet Large Scale Visual Recognition Challenge. The multi-layered CNN was able to recognize images of specific objects at a near-human level. After the 2012 success, the AlexNet research was sold to Google. To learn more about AlexNet, click here. -
Eugene Goostman passes the Turing Test
Eugene Goostman is not another computer scientist, but a chatbot. In a contest held by the University of Reading in 2014, the chatbot, who takes on the persona of a 13-year-old Ukrainian boy, held a convincing 5-minute, open-ended conversation that fooled 33% of judges to think it was a human. As a result, the chatbot passed the Turing Test, developed by Alan Turing in his 1950 paper, on the 60th anniversary of Turing’s death. -
OpenAI
OpenAI is leading the world in AI development. Beginning as a start-up in 2015, founded by Elon Musk, Sam Altman, and others, the mission of the company is “to ensure that artificial general intelligence benefits all of humanity.” Since 2015, OpenAI has released many AI programs to the public and has partnered with major companies, like Google, Facebook, and Microsoft. -
GPT-3
OpenAI released GPT-3 in 2020, and the program can create several genres of high-quality text from a prompt. GPT stands for “generative pre-training,” and this is OpenAI’s 3rd generation of the program. The program experienced quite a lot of criticism upon its release, but its ability to create text, or even code, in a human-like way with such ease shocked the world. -
DALL-E
Released by OpenAI in 2021, DALL-E is a text-to-image AI program that can create striking and unique images from written descriptions. DALL-E can create images from words in a prompt, or can expand on an existing image uploaded into the program. Watch a video of how to use DALL-E here. -
Minerva Math
Google announced the AI program Mivera in 2022. The program is able to solve college-level mathematic and scientific questions. Minerva is able to understand a question that uses both natural language and mathematic notation and solve it using step-by-step reasoning. While Minerva sometimes arrives at correct answers using incorrect reasoning, it is a step toward quantitative reasoning AI. -
Writing with ChatGPT
ChatGPT is OpenAI’s newest advancement in language AI. It acts in a conversational way to answer a prompt or create a textual form. ChatGPT recognizes patterns from internet data and past chat answers to create acceptable answer input for the user. The chatbot has excited and scared the world simultaneously. For a video about OpenAI and ChatGPT, click here.