Artificial intelligence (AI) tools are initially made to solve problems such as language assistance, legal affairs and crisis management.
Ayesha Khanna, the co-founder of an AI firm, discussed the solutions that these AI tools were made for at an information and technology fair in Hong Kong called InnoEx 2023 on April 12.
InnoEx 2023 opened with around 400 exhibitors showcasing their smart solutions to potential buyers, investors and entrepreneurs at the Hong Kong Convention and Exhibition Center (HKCEC) from April 12 to 15.
RELATED: ‘Smart’ Hong Kong: Robot mapping dog, traffic apps among new urban integrations |Robots vs COVID-19: Hong Kong used these machines to contain pandemic
Khanna was among the numerous industry experts who were invited to hold seminars at the fair.
She discussed the pros and cons of AI and AI-powered tools that are currently popular today.
“You don’t do AI for AI’s sake. You do AI because it solves problems,” Khanna said.
She also noted more than 500 cities in the world have also sought to build digital twin software or virtual copies that are paired with AI.
A digital twin is a virtual model or a simulation used to reflect a physical object.
Khanna cited traffic flow, rescue and crisis management and healthcare management as among the reasons for applying AI on a city scale.
Here are some of the widely adopted AI tools today and their purposes for their creation:
View this post on Instagram
Khanna is a co-founder of Addo AI, a global AI consulting firm. She was also named as a groundbreaking female entrepreneur in Southeast Asia by Forbes in 2017.
For language assistance
ChatGPT has gained at least 100 million users worldwide, including from the Philippines, since it was launched on November 30, 2022.
It is an AI-powered chatbot used to generate different types of texts at high speed.
It is developed by a company called Open AI in partnership with Microsoft.
RELATED: Microsoft introduces AI-powered cybersecurity assistant
On its website, the tool’s dialogue format was described to help “answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.”
In her talk, Khanna explained that ChatGPT was initially developed for language assistance.
She explained that it became particularly helpful for users who are not native English speakers.
To generate the texts, ChatGPT “calculates” or processes the best possible words based on the general knowledge it was fed into.
“Because it has all this knowledge, it is able to process it at high speed,” Khanna said.
The executive further emphasized that ChatGPT performs fast but “at every point in the sentence, it is calculating the probability.”
Another popular type of generative AI is a photo generator developed by a company called Stability AI.
Based on its website, two of its models, namely, Stable Diffusion and Deep Floyd IF generate images and photos through text prompts.
They produce unique images by processing all other types of pictures from the internet.
Khanna also noted that images produced by these tools are not made by any human photographer.
“It has looked at all the pictures in the world and it can generate a picture in seconds,” she said.
For legal affairs
A lesser-known AI tool named Harvey AI or Harvey is now being used by some law firms in the United States.
It assists lawyers with legal administrative tasks, thus lowering legal fees.
Khanna explained that Harvey was not developed to replace law graduates.
“This doesn’t [Harvey] will replace the lawyers. Those graduate students can now do other things. They can look at other patents related to AI. They can actually look beyond the groundwork of contract due diligence and actually think about strategy,” she said.
“This helps lawyers become more productive and bring down their price,” she added.
Harvey is also developed by Open AI in partnership with Microsoft.