Cloud Hosting

Financial institutions have rich, customer-centric data and are in a strong position when building AI solutions. Since the functions differ, AI use cases also differ.

The recent intense interest in generative AI has given rise to a new aspect of data science: Prompt Engineering, which is basically how humans train models like GPT by creating appropriate prompts.

At Data Summit 2023, Supreet Kaur,AVP,Morgan Stanley, discussed Leveraging Data Science and Generative AI during her session.

The annual Data Summit conference returned to Boston, May 10-11, 2023, with pre-conference workshops on May 9.

AI is transforming the quality of products, and such enhancements are making it possible to solve breakthrough problems and transform the day-to-day operations of companies in this area.

Drivers behind AI adaptation include big data, infrastructure, competition and automation, she explained.

Generative AI is an unsupervised and semi-supervised algorithm that enables computers to create new and original data that looks like humans generated it.

Youve all heard about GPT but, I feel that GANs arent getting enough attention, Kaur said.

Generative Adversarial Networks (GAN) helps identify misinformation, she noted.

Examples of generative AI use cases include financial and healthcare industries. In healthcare, generative AI can generate medical reports and clinical data based on conversations with doctors and provide support. Generative AI can provide personalized experiences to patients using their vitals, and generic conditions for preventative care.

Generative AI works by using prompt engineering, she said. Prompts are a set of instructions given to the model to generate the desired output. Prompt engineering is a natural language processing technique to create and fine tune prompts to get accurate responses from the model.

The future looks bright as we are going toward a low-code/no-code approach, Kaur said.

However, there are a few risks that generative AI brings. These risks include information leaks, bias, model hallucinations, ethical implications, and harm to society.

Data is the soul of any AI and ML algorithm, and hence spending massive amounts of resources and time can reap benefits in the future.

Choosing the suitable model is an inevitable step in your ML lifecycle as that will ultimately dictate the long-term success of your model, she said.

The industry is heading toward creating a more job roles in this area. This includes:

AI wont replace us, but a person using AI will, she said.

Many Data Summit 2023 presentations are available for review athttps://www.dbta.com/DataSummit/2023/Presentations.aspx.

View original post here:

Unlocking the Power of Data Science at Data Summit 2023 - Database Trends and Applications