Science and Tech

Artificial intelligence in banking is already a reality

Artificial intelligence takes over the search engine race

The artificial intelligence (AI) They are the systems and algorithms that machines create to solve problems in which human intelligence is required and in the banking operation it is already being implemented.

An example of this are the chatbot (assistant that communicates with users with text messages) and the AI ​​that generates images, that are causing a paradigm shift from the conventional conception of creativity and authenticity.

(Reach 40% of contact centers, Resolve’s goal).

According to asobancariaamong the AIs that generate images according to parameters specified by the user, it stands out DALL E 2developed by Open AI.

Regarding chatbots, ChatGPT It is today the AI ​​par excellence.

On November 30, OpenAI launched ChatGPTan AI language model that provides conversational responses generated based on user input.

(Why it would be a good thing if AI replaced humans in jobs).

The model combines reinforcement learning algorithms and human input that calibrate more than 150 billion parameters with which it answers follow-up questions, admits errors, questions, and rejects incorrect or inappropriate premises. The recent release of ChatGPT sparked interest in AI, because it can not only offer results as a machine learning model and it interacts in a human way with creativity and a certain emotion.

Asobancaria tried it out and by way of an interview asked him what his perspective is on the relationship between banking operations and AI.

asobancaria: “What will the reality of the banking and financial sector be like in a few years with the increasing appearance of AI?

ChatGPT: “In the near future, AI and other technologies are likely to have a significant impact, transforming the way transactions are conducted and customers are served. Automation and personalization are expected to increase efficiencies and improve the customer experience and transform the financial industry”.

(Supercomputer predicted what the end of the world would be like: this said).

asobancaria: “Will it be possible to integrate tools like ChatGPT into financial sector operations efficiently?”.

ChatGPT: “Yes. Tools like ChatGPT are likely to be integrated into financial sector operations in the near future. AI and natural language processing, like the one used by ChatGPT, can be useful for automating repetitive tasks, improving efficiency and delivering personalized services.”.

asobancaria: “What could be the challenges of implementing AI in the sector?”.

ChatGPT: “There are various challenges including data quality, regulations and compliance, job disruption, trust and accountability, and integration with legacy systems.”.

‘Perfect’ images with AI

According to Asobancaria, progress in AI is not limited to conversational interactions. Turn text descriptions into images with programs like Midjourney, the open source Stable Diffusion, and DALL E of OpenAI are a sample of its evolution.

In January 2022, OpenAI introduced DALL E 2 (its latest system), which generates more realistic and accurate images. It can create art and images from a text-type description that combines concepts, attributes and styles, make edits to existing images from a natural language caption, add and remove elements taking into account shadows, reflections, textures, etc.

(‘My worst fear is causing significant damage’: creator of ChatGPT).

As with ChatGPT, the authors of the model seek to curb misuse by limiting the ability to generate violent, political, hateful, or adult images, removing explicit content from training data, minimizing exposure to these concepts, and using advanced techniques to avoid photorealistic rendering.

BRIEFCASE

Source link