Regulating Artificial Intelligence

8

October

2016

No ratings yet.

Artificial Intelligence, a term originally coined in the 1950s, received increased attention within the last year due to technical progress within machine learning and related technologies and an increased number of practical applications building on it.

 

This technical development has led to many ethical concerns being raised.

Some are derived from current issues within software development like life/death decisions made by autonomous vehicles, the adoption of stereotypes through biased training data or the treat of rising unemployment through automatization.

Other raised concerns are more theoretical or futuristic in their nature and deal with scenarios of machines taking over as displayed in science fiction movies like Terminator or the Matrix Trilogy. Another well discussed thought is the “technical singularity” a term describing a future after artificial intelligence reaches above human level  cognitive capabilities.

 

No matter if for current or future issues, there seems to be a need for guidelines or rules that deal with these questions. Few political voices have raised the topic already leading to the government of the United States investigating the topic for potential needs for regulation at the moment.

 

But private entities are addressing these moral concerns as well.

 

Elon Musk, CEO of Tesla and Space X, and Sam Altman of Y Combinator recently found the  non-profit artificial intelligence research company OpenAI in order to “build safe AI, and ensure AI’s benefits are as widely and evenly distributed as possible”. Since the launch in April they have been publishing their work in the shape of research articles and blog posts and released OpenAI Gym, a toolkit for developing and comparing reinforcement learning algorithms. The long term of the non-profit is to develop safe and open-source algorithms that lower the risk of disasters through mistakes in AI creation and the prevention of monopolies within knowledge distribution to certain big companies.

 

Exactly these big companies however recently came together to create the “Partnership on AI”. Facebook, Amazon, Google, IBM and Microsoft, in an apparent act of self-governance, formed the group to exchange knowledge, create best practices and publish research especially in the fields of ethics, inclusivity and privacy. To prevent concerns or conspiracy theories, the partnership plans to make discussions and minutes from meetings publicly available.

 

It is currently too early to judge the output created by these organizations, but it certainly will be interesting to keep an eye on them and track the developments leading to human-friendly decisions and algorithms or to potential problems in the future.

 

Sources:

US Government

https://www.whitehouse.gov/blog/2016/05/03/preparing-future-artificial-intelligence

 

Open AI

https://openai.com/

https://www.wired.com/2016/04/openai-elon-musk-sam-altman-plan-to-set-artificial-intelligence-free/

 

Partnership on AI

http://www.partnershiponai.org/

https://techcrunch.com/2016/09/28/facebook-amazon-google-ibm-and-microsoft-come-together-to-create-historic-partnership-on-ai/

Please rate this

Algorithm marketplaces

26

September

2016

No ratings yet.

Algorithms, the “DNA of software,” have become central to how our digitized world works. From what you see in your Facebook news feed, to which traffic lights are green on your way to work, to what discount you may get when buying groceries in the supermarket – it is all defined through invisible software processes.

In the last decade we have seen both raw data and information increasingly being sold as goods or made accessible for a subscription fee. Algorithms however are currently predominantly developed and reused directly by their owners, published in academic journals or distributed as open source code.

Some entrepreneurs however, recently came up with the idea of algorithm marketplaces. These work like app stores for software building blocks or, as the MIT Technology Review puts it, “dating site[s] for algorithms.” These platforms connect academics or software developers, who want to monetize their creations, with others who are trying to find existing solutions to problems they are facing.

Forbes lists the emergence as an important prediction for the future of Big Data while Gartner sees substantial growth opportunities and predicts that “the three leading marketplaces will offer more functions than all other analytical vendors combined – transforming the analytics and broader software market in the process.”

Diego Oppenheimer, a former program manager at Microsoft, and Kenny Daniel, who did his PhD in artificial intelligence at the University of Southern California, are now on the mission of freeing the algorithms currently “trapped in white papers.” Their company Algorithmia is one example for an algorithm marketplace, providing a cloud platform onto which users can upload their algorithms and make them accessible through APIs. To access them, users pay an amount defined by the developer (30% of which goes to the platform) and are charged for computational power. Apart from the listing of existing algorithms, Algorithmia provides a “bounty system”, which lets users ask for special algorithms that can then be developed from scratch or be reused by others.

There is also a variety of other marketplaces for algorithms for specific fields like ad placement (DataXu) or machine learning (Algorithms.io), but Algorithmia is the first one that offers and sells any type of algorithm.

Gartner notes that algorithm marketplaces could contribute to an even further growth within the software market by creating a new niche for software snippets both reused and developed particularly for these platforms. In addition, they could allow the development of more complex software, allowing developers to build on existing algorithms.

The founders hope that this platform will lead to a democratization of software development, especially in the area of deep learning systems where it currently seems like few software giants are about to centralize talent and knowledge.

Sources
17 Predictions About The Future Of Big Data Everyone Should Read
http://www.forbes.com/sites/bernardmarr/2016/03/15/17-predictions-about-the-future-of-big-data-everyone-should-read/#47b6f89a157c

A Dating Site for Algorithms
https://www.technologyreview.com/s/530406/a-dating-site-for-algorithms/

Algorithmia Raises $2.4M To Connect Academia And App Developers
Algorithmia Raises $2.4M To Connect Academia And App Developers

Algorithms move from academia to marketplace

Algorithms move from academia to marketplace

The Algorithm Economy Will Start a Huge Wave of Innovation
http://www.gartner.com/smarterwithgartner/the-algorithm-economy-will-start-a-huge-wave-of-innovation/

Wanna Build Your Own Google? Visit the App Store for Algorithms

Wanna Build Your Own Google? Visit the App Store for Algorithms

Please rate this