Artificial Intelligence Assistant

From SI410
Revision as of 03:57, 9 February 2023 by Hayleyc (Talk | contribs)

Jump to: navigation, search

Artificial intelligence assistants are comprised of a software program or system that aims to mimic human intelligence and perform daily tasks that typically would require human-level cognition. These AI-powered assistants are capable of learning from user interactions, adapting to new situations, and pulling context and data from user's metadata. By analyzing the user's previous patterns of interaction, it allows for personalized responses catered to the user's needs. These assistants rely on technologies like NLP (natural language processing), ML (machine learning), and data science to be able to provide accurate data and perform tasks correctly. They are able to take on a wide variety of tasks, such as answering basic questions, scheduling appointments and alarms, and controlling smart home devices. Because of the increasing versatility and capabilities of AI-powered assistants, they are rapidly growing in the market and becoming more prevalent in the day to day lives of many. Artificial intelligence assistants are commonly implemented and categorized into chatbots, virtual personal assistants, conversational agents, and intelligent home devices.

History

Trial and Error: 1960s-1990s

The history of artificial intelligence assistants traces back to the 1960s, where the first ever chatbot, ELIZA, was developed by MIT professor and researcher Joseph Weizenbaum.[1] ELIZA worked on pattern matching and had scripted responses that were given when faced with a problem statement. It passed the words inputted by the user to a computer, and then paired those words to a list of possible scripted responses. Weizenbaum designed ELIZA to simulate a therapist, such that it accepts user's questions and statements, and returns a response. Although this model's capabilities were quite limited, it still marked the beginning of the development of artificial intelligence assistants.[2] Many chatbot makers built upon Weizenbaum's model to strive for more human-like interactions between chatbot and user.[3]

Gaining Success and Popularity: 2000s-present

As AI research continued to advance, the development of natural language processing (NLP) and machine learning (ML) enabled AI to expand its functionalities. The 21st century marks the emergence of conversational AI, which further led to much more advanced technologies. SmarterChild, an intelligent bot developed by ActiveBuddy in 2001, was in many ways a true precursor to Siri.[4] It had the ability to carry out fun conversations with the user, as well as have access to data.[5]

In the early 2010s, several companies announced their first virtual personal assistants (VPA), such as Apple's Siri, Amazon's Alexa, and Google Assistant. These AI-powered assistants were designed to handle advanced tasks such as answering questions on command and scheduling tasks and appointments. They were also integrated with other smart devices, such as smartphones, speakers, and home devices, making them much more versatile and easily accessible.

In 2023, the use of AI assistants in everyday life has become much more common. New technologies are allowing for the further advancement of AI and AI products. These AI systems include chatbots like ChatGPT from OpenAI, virtual reality systems in the Metaverse, and social media systems in Tiktok and Instagram.[6] ChatGPT is the most advanced chatbot to present day, with the ability to carry out a variety of tasks ranging from conversation generation to language translation. [7]

Functionalities and Methods of Interaction

Platforms and Services

Virtual Assistants

Some of the most widely known examples are virtual assistants, such as Apple's Siri and Google Assistant. Both Siri and Google Assistant use natural language processing and voice queries to function.

Smart Speakers

Another category of products that have become very popular in recent years is smart speakers. Amazon Echo, Google Home, and Apple HomePod [8] are all examples of successful smart speakers that implement AI technology to allow for an interactive experience. These speakers typically focus on controlling smart home devices, playing music, and answering questions.

Chatbots

Although chatbots have been around since the 1900s, they have been constantly adapting and advancing with technology. ChatGPT is a perfect example of how chatbots have evolved and become highly used in today's society. It is a language model that is capable of understanding and generating text for any given context. The software uses deep learning and was trained on a large corpus of text data, enabling it to produce answers that are of great precision, detail, and coherence.[9]

Benefits and Societal Impact

Through the advanced technology and functionality of AI assistants, it brings several benefits to society and industries. A major benefit that can evidently be seen is increased productivity, in both personal lives and in organizations. Their ability to predict and automate tasks that would typically be repetitive allows for users to allocate their time in a much more efficient manner. This allows users to focus and spend more of their time completing complex and creative tasks. AI assistants also provide quick access to useful information, such as news, daily weather, and directions to a given location. This streamlines the task of having to manually check this information on a user's phone or laptop, increasing productivity.

The impact that AI-powered assistants have had on industries and businesses are also evident throughout the years. Their ability to automate basic but time-consuming tasks have increased efficiency and productivity across a wide range of industries. Businesses have sought multiple call center solutions, and using intelligent virtual agents have proven to be useful. They can handle a large volume of customer inquiries, thus freeing up human agents to focus on helping those who need personalized assistance. In addition, due to their automated nature, they are available around the clock and can provide service to customers at any time of the day. [10] This results in faster response times, improved customer satisfaction, and allows for businesses to allocate their spendings elsewhere. Another benefit that AI assistants have on industries is their ability to process and analyze data, as well as make decisions based on this data. By processing these large amounts of data in real-time, they can provide businesses with valuable insights and recommendations. For example, in the retail industry, AI-powered assistants have the ability to analyze sales data to identify trends, and make recommendations for marketing strategies and inventory management.[11]

Another aspect of AI assistants that typically have a positive outcome is the personalization factor. Through ML and analyzing previous patterns, an AI assistant is able to adapt to the needs of each user. This presents extremely personalized responses, which helps to increase user engagement and improve customer satisfaction.[12] A successful example of this is ChatGPT, which utilizes a user's previous questions and responses to generate more accurate results. This concept of personalization highly attracts users, allowing for a smooth conversational and dynamic interaction between the bot and user.[13] Although this is usually seen as a benefit, if presented in the wrong way it could lead to potential privacy and ethical concerns.

Ethical Implications

Privacy

A major concern among users is privacy. As these assistants are designed and programmed to collect and store a large amount of personal information, there is a risk that this information could be accessed or misused by unauthorized parties. Because of their access to personal data such as location, purchasing habits, search history, and communication, it raises the question of how this data is being stored and used. There are concerns that this data is being used to generate a user's profile, and make customized decisions such as showing personalized ads to the user. Additionally, the constant surveillance and tracking of users by AI assistants creates concerns about the potential for the erosion of personal privacy and autonomy. [14] Especially for voice-activated assistants, such as smart speakers, users have voiced a major concern about "who is listening" on the other side. Users worry that this audio data could potentially be misused or recorded in a manner that they did not consent to. [15] This places an emphasis on businesses to consider the importance of protecting and being transparent about how the data collected from the user is being used. Security mechanisms such as end-to-end encryption, two-factor authentication and biometrics are a few possible solutions to add a level of security to an AI-powered assistant. [16]

Biases and Discrimination

Another ethical concern raised by AI assistants is their potential to perpetuate biases and discrimination. As AI assistants are often trained on data sets that reflect the biases of their creators, there is a risk that they will perpetuate these biases in their responses and recommendations. This can have a particularly harmful impact on marginalized groups, such as people of color and members of the LGBTQ+ community. In addition, there is a strong gender bias when it comes to the voices of AI assistants. Alexa, Siri, Cortana, and Google assistant all share one similar feature - the default voices are female. This can reaffirm the social construct and gender bias that women are deemed as subservient, polite, and should be the ones to "assist" others.[17] This poses an issue particularly for those who use AI assistants in a family setting. For children who are growing up using these assistants, they may easily come to a conclusion that females should always be the ones helping and assisting others. Another issue that is raised by AI assistants, specifically those that are voice-powered, is that some users may have accents or a different dialect of the language that the AI machine was programmed to understand. If the AI assistant was not trained to understand these different dialects, it could create a sense of discrimination against those who do not speak the language in a "perfect manner". [18]

References

  1. https://onlim.com/en/the-history-of-chatbots/
  2. https://yellow.ai/chatbots/evolution-of-virtual-assistants/
  3. https://onlim.com/en/the-history-of-chatbots/
  4. https://www.vice.com/en/article/jpgpey/a-history-of-smarterchild
  5. https://onlim.com/en/the-history-of-chatbots/
  6. https://www.nytimes.com/2022/12/29/technology/personaltech/new-tech-2023-ai-chat-vr.html
  7. https://onlim.com/en/the-history-of-chatbots/
  8. https://techresearchonline.com/blog/best-ai-assistant-of-2022/
  9. https://www.assemblyai.com/blog/how-chatgpt-actually-works/
  10. https://getvoip.com/blog/intelligent-virtual-assistant/
  11. https://www.druidai.com/blog/5-benefits-of-ai-powered-virtual-assistants-across-the-retail-industry#:~:text=AI%2DPowered%20Virtual%20Assistants%20provide,offer%20personalized%20and%20convenient%20support.&text=Conversational%20AI%20in%20retail%20has%20dramatically%20changed%20the%20industry.
  12. https://mobidev.biz/blog/ai-virtual-assistant-technology-guide#:~:text=Siri%2C%20Cortana%2C%20and%20Google%20Assistant,developers%20integrate%20into%20their%20applications.
  13. https://www.atriainnovation.com/en/how-does-chat-gpt-work/
  14. https://aisera.com/chatbots-virtual-assistants-conversational-ai/
  15. https://www.privacypolicies.com/blog/voice-assistants-privacy-issues/
  16. https://mobidev.biz/blog/ai-virtual-assistant-technology-guide#:~:text=Siri%2C%20Cortana%2C%20and%20Google%20Assistant,developers%20integrate%20into%20their%20applications.
  17. https://www.adaptworldwide.com/insights/2021/gender-bias-in-ai-why-voice-assistants-are-female
  18. https://aisera.com/chatbots-virtual-assistants-conversational-ai/