We’re all aware of the rise of new AI technologies like ChatGPT, Dall-e, and Midjourney. These breakthroughs have spawned a plethora of fresh uses and concepts. However, the question for small and medium-sized businesses remains: how can they take advantage of this technological boom?
Using AI to fuel customer service chatbots, automating tedious operations, and boosting decision-making through predictive analytics are a few examples. However, there are plenty more alternatives worth investigating.
In this article, I will offer you with full instructions and examples of how you may integrate AI in your company right now without having substantial technical knowledge or expensive equipment. I’ll go over some of the more intriguing concepts I’ve come across in recent weeks. There will be instances of various technical challenges, ranging from “copy and paste” into ChatGPT through the deployment of AzureAI services and virtual devices.
Smart tutor
Learning is a very personal experience, with each individual taking their own way of assimilating new information. AI can be a valuable tool for businesses that require ongoing training or want to support their new employees’ onboarding process. AI can help people learn more effectively and efficiently by giving a self-paced and personalised experiences.
Using AI models to construct teaching courses is undoubtedly one of the most inventive techniques that businesses can adopt. Companies can fine-tune the AI model on specific themes to deliver a more tailored learning experience by using a RAG [Retrieval-Augmented Generation] architectural model, as recommended by Tobias Zwingmann in his article, or Azure OpenAI services to limit the data and boost the governance.
JushBJJ’s GitHub repository contains an example of this tutor technique, which includes a JSON prompt that turns the ChatGPT online model into a personalised instructor. This tutor can explain subjects linked to the data provided as well as test the employee’s knowledge and responses, which is very beneficial for continuing certifications.
Here are some links to relevant GitHub repositories:
- RAG Architecture Model: [link here] – Expert Level
- Own infrastructure (or cloud)
- ChatGPT API
- App Development
- Azure Open AI model: [link here] – Expert Level
- Access to Azure Cloud
- App Development
- JushBJJ’s Mr.-Ranedeer-AI-Tutor – Copy/Paste Level
- Just sign to ChatGPT
- Download JSON file and change values if needed
- Paste on the web
Data Storytelling
The contrast between data analytics and data storytelling has been a source of contention on the internet. It can be difficult to find people who are good at both, but AI is good at summarising and assessing data. With a little help, this can be easily improved and turned into an excellent data story.
One method to use AI in this scenario is to prompt ChatGPT with your data. You may then use AI to assist with data summarising and storytelling. But we don’t want to manually replicate every single row of our data into ChatGPT. So we need connect our data analytics tool to the ChatGPT API.
Mathias Halkjaer’s blog walks us through the process of connecting ChatGPT API to our PowerBI data. This method can be used to summarise raw data and then easily extended to give more appealing narrative based on other criteria present in the data.
- ChatGPT and PowerBI: [link here] – Level Intermediate
- Requires ChatGPT API
- PowerBI Desktop
- PowerBI Development knowledge
Fast cyber defense
AI is increasingly being used in cyber security, where hackers are already using AI technologies to gain access to networks and attack vulnerable systems. This time we can try to turn the tables and use artificial intelligence to protect ourselves against cyber threats.
In his blog post, Xavier Bellekens suggested a fresh approach, demonstrating how ChatGPT’s coding capabilities may be used to quickly create a network honeypot. This technique enables security experts to detect breaches by offering an easy target that hackers could try to exploit to elevate privileges or establish their presence on the target network. However, this target is not the real expected device and does not have any capabilities to cause harm, making the attackers waste precious time and resources.
- Deceiving Adversaries – Level Intermediate
- a VM or network device (Raspberry Pi or similar) capable of running Python
- Some python knowledge to modify the provided output from ChatGPT and tweak as needed
- Network and security knowledge
Other well known uses of AI for cyber defense that are present in several commercial packages and offerings include:
- Behavior-based threat detection and prevention
- Streamlining security operations
- Rapid incident response and mitigation
Conclusion
Businesses that employ AI technology will be far more prepared to succeed in the future as the technology evolves. Companies may use AI to streamline their processes, make data-driven decisions, and strengthen their security defences against cyber attacks. AI, with its expanding capabilities and possible uses, is a vital tool that may help businesses prosper in today’s digital world.
