Keshav Mittal,
Editorial Intern,
Indian Society of Artificial Intelligence and Law.
In the year 2020, we have achieved a quite a few milestones when it comes to artificial intelligence which is as astonishing as they are horrifying, having undoubtedly enormous power which can be used or misused in ways we can not even imagine.
To start with the latest GPT - 3 Generative Pre-trained Transformer 3, it is basically a text generating tool created by OpenAI. “Seeded with a few sentences, like the beginning of a news story, GPT-3 can generate impressively accurate text matching the style and content of the initial few lines — even down to making up fabricated quotes.”
It is the sequel of GPT - 2 which was titled as the world's most dangerous algorithm and GPT - 3 even supersede the potential and power of it. This technology is considered dangerous because it can be misused by many organisations to spread wrong messages about miss believes like Jihad, racism etc. It will also make it harder for the people to distinguish between the actual text and the text created by artificial intelligence. GPT - 3 has 175 billion parameters which are used to make the text more brain-like.
OpenAI is not the only company developing technologies in this field, many more companies like Microsoft and Google have also launched their versions of this technology like Ok Google's transformer, 2017 and Microsoft Turing Natural Language Generation T - NLG, 2020. This software are undoubtedly very inferior to GPT - 3, in terms of both power and technology as T - NLG has only 17 billion parameters in comparison to GPT- 3's 175 billion parameters. The cost incurred in developing these Technologies is also very huge as it cost almost $ 1 per 1,000 parameters to train a network.
As dangerous as it seems artificial intelligence is also been and will be used for the good of mankind. Many technologies are been developed in the field of Medical Science to make treatment of more and more diseases possible and more effective. A start-up called SpeakUnique is trying to develop new technology which can convert your thoughts into speech for disabled people. This is a very ambitious but great initiative which can help many people to convey their thoughts. This technology works by translating a person's brain signals into the synthesized speech of that same word or sentence.
There are also other uses of A.I. like to predict the shape of protein-based on the sequence. This technology would Revolutionary in biotechnology as proteins are the building blocks of the human body and it is necessary for us to understand it as good as we can. This technology can predict the shape of a protein in our body all the basis of their sequence.
There have been many theories about technology taking human jobs and not even theories there are many real-life examples of AI-based Technology replacing humans at the workplace. In the year 2020 many jobs have been lost all over the world but not because of the technology replacing humans rather due to pandemic and it's After Effects. What are ionic is that the most employment given in this year was by big Tech Giants hiring people for developing these technologies. So in short term, there's no fear of such thing as Technology replacing humans but in a few years, it is a very strong possibility.
Deep fake is one of the most misused Technology in the world. It is used for altering a person's face or voice in a photo or a video with other person's face or voice typically used to spread false information. There is a recent incident in which a high budget deep fake video was made by the “researchers from the Center for Advanced Virtuality at the Massachusetts Institute of Technology.” The video was showing President Richard Nixon giving a speech about Apollo 11 moon landing mission by NASA. It was an alternative speech which was prepared just in case to be given if the mission would have failed. So this technology is very dangerous and can be used for far more malicious purposes and can cause a greater threat than this.
This calls for the rules and regulations to protect us from the malicious use of artificial intelligence and its Technologies. For example, California has passed a legislation AB - 730. “This law designed to criminalize the use of deepfakes to give false impressions of politician’s words or actions.” There are many more tech laws but due to the quick developments, there is always something or the other left or some loophole left to be exploited. So to prevent that from happening we have to update our tech laws regularly and there should be governmental supervision over the development of these kinds of dangerous technologies.
Artificial intelligence is getting advanced day by day and new and far superior Technologies will come every day. These technologies are very good for human society will help us develop grow and achieve greater Heights but this will come with its fair share of threats and dangers like digital terrorism, privacy violations, social-economic inequality, automation spurred job loss, deepfakes, weapons automatization, digital warfare etc. To prevent these things there is a need for strong and strict laws for the use and protection of artificial intelligence. It should be protected from getting in the hands of peoples in conflict with the law and want to use this technology for some ulterior and malicious.
Bibliography
1. “Dormehl Luke A.I. hit some major milestones in 2020. Here’s a recap. December 4, 2020. https://www.digitaltrends.com/features/2020-ai-major-milestones/?utm_campaign=78055_Editorial_Decrypted_12042020&utm_medium=email&utm_source=dotdigital&dm_i=6MG2,1O87,13CZNZ,7OE0,1
2. Dormehl Luke Amazing new headset translates thoughts into speech for vocally impaired wearers November 5 2020. https://www.digitaltrends.com/features/synthetic-voice-gaming-headset/
3. Griffin Andrew, AI DEEMED ‘TOO DANGEROUS TO RELEASE’ MAKES IT OUT INTO THE WORLD, 7 November 2019 https://www.independent.co.uk/life-style/gadgets-and-tech/news/ai-artificial-intelligence-dangerous-text-gpt-2-elon-musk-a9192121.html
4. Dormehl Luke, An alternate history of the moon landing, brought to you by deepfakes. July 20, 2020 https://www.digitaltrends.com/features/mit-alternate-history-of-moon-landing-deepfake/
5. Dormehl Luke, OpenAI’s GPT-3 algorithm is here, and it’s freakishly good at sounding human. 12 June 2020, https://www.digitaltrends.com/features/openai-gpt-3-text-generation-ai/”