The idea of intelligent machines has long fascinated science fiction writers, but in recent years, they are increasingly becoming part of reality.
IBM’s Watson, an artificially intelligent computer system that can answer questions and understand natural language, is an indication of how sophisticated the technology has become.
In 2011, Watson won the US quiz show Jeopardy! against human competitors and since last year it has been used by a hospital in Germany to diagnose rare diseases. Watson is also being used by a Japanese insurance firm to to calculate payouts to 34 job losses.
Watson is an example of artificial intelligence (AI), the concept of machines that can mimic the sophisticated capacities of the human brain. While we are a long way from having a computer system that can think just like us AI is already widely used.
For example, machine learning where a computer system is trained to identify patterns in data, and then use the knowledge it has gained to make predictions, is used in a number of different sectors. It powers the auto-complete feature on internet search engines, which then use information from past searches to predict what people are looking for. Other applications include identifying banking fraud and steering self-driving cars.
Machine learning could potentially be used by museums to improve performance in areas such as marketing, fundraising and operations. A project last year by the innovation charity Nesta used it to predict the rough size of museum visitor numbers using check-in data from Foursquare, a social network.
John Davies, a research fellow at Nesta who worked on the project, says that machine learning has many potential uses for museums, such as helping museum professionals understand collections, improve digital records and identify art forgeries.
“If museums have very large collections that have been digitised, in principle there is scope for running these kinds of pattern recognition algorithms on the collection, and understand it better in a way that would be difficult to do manually,” Davies says.
Developers are already able to call on a number of established image recognition AI tools based on machine learning, or its more sophisticated relative, deep learning. But Mario Klingemann, an artist in residence at the Google Cultural Institute, says that the question of how these tools can best be employed to benefit museums, curators and visitors is a live one.
“What ways are there to make work easier, discovery easier and help with research? There are tons of different roads you can take,” Klingemann says.
The Google Cultural Institute has a microsite showcasing experimental projects that aim to use the many thousands of digitised images to create new experiences and increase access to art.
In t-SNE map, visitors can explore a 3-D landscape of artworks organised by visual similarity, while in Curator table users can ask the programme to arrange artworks under different categories such as artist, subject, colour and chronology. And in Klingemann’s X Degrees of Separation, the system generates a visual pathway of artefacts that link any two chosen by the visitor.
But even at this experimental stage, the idea of ceding curatorial control to a machine throws up a number of ethical questions that are likely to become increasingly important as AI develops.
One project that aimed to get the public thinking critically about AI was Recognition, which won Tate’s IK Prize last year. A website and an exhibition at Tate Britain were used to compare contemporary photojournalism to British art collection using AI algorithms.
“We thought it would be interesting to apply rational, objective, computational thinking to a subjective field like art. Combining these two entities could create a new way of looking at the Tate collection,” says Angelo Semeraro, an interaction designer at the Italian communications research centre Fabrica, which created the project.
Of course, it isn’t just the public who are unfamiliar with AI. Andrew Larking, the creative director of digital agency Deeson, who has a background in the museum sector, says that there is also a lack of understanding in museums about what AI is and what it can do.
“I speak regularly to people who work at museums, art galleries, archives, and libraries, and the perceptions of what AI is and how it can be used come across like they've been ripped out of a 1950's B movie about robot invasions,” Larking says.
He believes that the confusion arises because AI is a concept rather than a tangible product, and adds it will be years before the sector is fully reaping the benefits of the technology.
“I would guess that it's going to take a decade for the understanding of what AI is and what you can do with it to find its way into the museum sector,” Larking says.
The MA's event Museum Tech 2017: A Digital Festival for Museums, taking place at the Museum of London on 29 June, will explore new and emerging digital technologies that are helping to shape the way audiences experience museums and their collections.
IBM’s Watson, an artificially intelligent computer system that can answer questions and understand natural language, is an indication of how sophisticated the technology has become.
In 2011, Watson won the US quiz show Jeopardy! against human competitors and since last year it has been used by a hospital in Germany to diagnose rare diseases. Watson is also being used by a Japanese insurance firm to to calculate payouts to 34 job losses.
Watson is an example of artificial intelligence (AI), the concept of machines that can mimic the sophisticated capacities of the human brain. While we are a long way from having a computer system that can think just like us AI is already widely used.
For example, machine learning where a computer system is trained to identify patterns in data, and then use the knowledge it has gained to make predictions, is used in a number of different sectors. It powers the auto-complete feature on internet search engines, which then use information from past searches to predict what people are looking for. Other applications include identifying banking fraud and steering self-driving cars.
Machine learning could potentially be used by museums to improve performance in areas such as marketing, fundraising and operations. A project last year by the innovation charity Nesta used it to predict the rough size of museum visitor numbers using check-in data from Foursquare, a social network.
John Davies, a research fellow at Nesta who worked on the project, says that machine learning has many potential uses for museums, such as helping museum professionals understand collections, improve digital records and identify art forgeries.
“If museums have very large collections that have been digitised, in principle there is scope for running these kinds of pattern recognition algorithms on the collection, and understand it better in a way that would be difficult to do manually,” Davies says.
Developers are already able to call on a number of established image recognition AI tools based on machine learning, or its more sophisticated relative, deep learning. But Mario Klingemann, an artist in residence at the Google Cultural Institute, says that the question of how these tools can best be employed to benefit museums, curators and visitors is a live one.
“What ways are there to make work easier, discovery easier and help with research? There are tons of different roads you can take,” Klingemann says.
The Google Cultural Institute has a microsite showcasing experimental projects that aim to use the many thousands of digitised images to create new experiences and increase access to art.
In t-SNE map, visitors can explore a 3-D landscape of artworks organised by visual similarity, while in Curator table users can ask the programme to arrange artworks under different categories such as artist, subject, colour and chronology. And in Klingemann’s X Degrees of Separation, the system generates a visual pathway of artefacts that link any two chosen by the visitor.
But even at this experimental stage, the idea of ceding curatorial control to a machine throws up a number of ethical questions that are likely to become increasingly important as AI develops.
One project that aimed to get the public thinking critically about AI was Recognition, which won Tate’s IK Prize last year. A website and an exhibition at Tate Britain were used to compare contemporary photojournalism to British art collection using AI algorithms.
“We thought it would be interesting to apply rational, objective, computational thinking to a subjective field like art. Combining these two entities could create a new way of looking at the Tate collection,” says Angelo Semeraro, an interaction designer at the Italian communications research centre Fabrica, which created the project.
Of course, it isn’t just the public who are unfamiliar with AI. Andrew Larking, the creative director of digital agency Deeson, who has a background in the museum sector, says that there is also a lack of understanding in museums about what AI is and what it can do.
“I speak regularly to people who work at museums, art galleries, archives, and libraries, and the perceptions of what AI is and how it can be used come across like they've been ripped out of a 1950's B movie about robot invasions,” Larking says.
He believes that the confusion arises because AI is a concept rather than a tangible product, and adds it will be years before the sector is fully reaping the benefits of the technology.
“I would guess that it's going to take a decade for the understanding of what AI is and what you can do with it to find its way into the museum sector,” Larking says.
The MA's event Museum Tech 2017: A Digital Festival for Museums, taking place at the Museum of London on 29 June, will explore new and emerging digital technologies that are helping to shape the way audiences experience museums and their collections.