Reading Time: 4 minutes
Nick McDonald, Account Manager at Fujitsu takes a closer view of the revolutionary technology, and says that it shouldn’t be feared when used correctly.
AI dominates the conversation in the industry today.
Businesses know that it has a profound impact on their entire operation, from the internal processes and procedures, to the tools, solutions and services provided to customers.
While there has been much hype about the potential of AI, concerns have also been raised.
Many organisations, and indeed the global market, are now apprehensive about AI, and uncertain of how to approach it.
There are many different types of AI, each with its own set of applications. This includes, but is not restricted to:
- Narrow AI
- General AI
- Super AI
- Reactive machines
- Memory Limits
- The theory of mind
- Self-awareness
Narrow AI, and the Generative AI within it, is the category of AI most people are interested in.
What do we mean by Generative AI in reality?
Generative AI – a simple definition:
The AI that generates new data can include text, images, video and more. This is done by learning from the patterns and structures of its training data in order to create new ideas that have similar characteristics.
This can include training in human language, programming, art, chemistry and biology, or law.
Generative AI is powered by large AI models. They are often referred to foundation models and can perform many tasks, including summarizations, classifications, and answering questions.
It is ideal for chatbots, media assets, product design, and other areas.
ChatGPT catapulted Generative AI to the mainstream:
ChatGPT, the most popular example of Generative AI, is used by many.
The rise of Instagram has been meteoric. It reached 100 million users within just two months. Facebook took 54 months to achieve the same milestone.
ChatGPT, a chatbot, is built on large language models. These LLMs are trained using vast amounts of data and produce text that can be understood by humans.
ChatGPT analyzes the meaning of the question asked by the user and determines what it is that the user wants to achieve.
Then, it returns the words and sentences that it predicts it will use to answer the query based on data it was trained on.
The risks of Generative AI:
ChatGPT is a fun tool, but it does highlight some concerns about Generative AI.
It can be a matter of accuracy, inconsistency, bias, lack of explanation, or even threats to privacy, security and intellectual property.
One of my favorites is the lawyer who used ChatGPT to look for precedents on a case he was working on.
The lawyer entered their question into ChatGPT, but luckily did not include any specific details about the case. It returned information that showed there were precedents in law.
Then they instructed ChatGPT compile all the information into a document that would be submitted to the court.
The judge found that the information ChatGPT provided was incorrect after reading the document presented by the attorney and doing their own research on the legal precedents.
The judge asked the lawyer to explain why he had used ChatGPT for searching precedents but not checked the results.
The lawyer was eventually struck off. This is the best result possible, because if they had included information about the case during their ChatGPT searches, they could have been sentenced to prison for violating confidentiality.
ChatGPT, as a public service, both consumes and shares information.
Generative AI is not a threat:
This shouldn’t deter organisations from harnessing the power of Generative AI, and reaping the benefits that it can bring.
Online gambling businesses can grow by improving their customer service, streamlining the development of games and streamlining their game development.
We have created a chatbot that is similar to ChatGPT, which organisations can use confidently. It is similar to ChatGPT, but it uses the data from the company’s warehouse instead of publicly available data.
It allows employees and teams to ask questions, and receive answers based on proprietary data fed into the foundation models.
The data can be separated into silos within the warehouse. Individuals and teams will have different levels of access. For example, the marketing team does not need to be able ask the same questions about the data as the CEO and CFO.
There are many uses for a private GPT. These include the ability to ask questions, chat with data from a company, discover new insights and answers, encourage deeper collaboration, generate insight jointly, and evaluate results.
This is only one example of how Generative AI could be used, but it shows that its capabilities can safely be harnessed without negative consequences.
As the industry learns to use the new technology, more and more products will be developed that will undoubtedly transform the game for everyone.
