From Concerned to Optimistic: What an AI Strategist Learned from Generative AI with Large Language Models

Generative AI (GenAI) is poised to transform work, but what that looks like remains unclear at the moment. While the International Monetary Fund believes 60 percent of jobs may be impacted by AI, other organizations estimate that the number could be closer to 70 percent. 

Knowing how to use AI as a tool and developing key skills around this technology will be critical in the coming years. Yet, as of today, less than 30% of professionals use GenAI at work on a daily or weekly basis. 

With AI technologies evolving at a rapid clip, it’s important to hear how professionals use it the world over. Coursera spoke with Mohammad, a Senior AI Strategist at a German consulting firm, who was initially both excited and skeptical  about the performance of large language models (LLMs). 

 Mohammad worried about how he could use LLMs reliably in business because they seemed particularly volatile with outputs that could vary widely. He also feared that only big tech companies could afford to develop the most powerful models, potentially leaving smaller players behind. 

However, in an effort to learn more about this rapidly advancing technology and dive deeper into the technical aspects of GenAI, Mohammad turned to Coursera and enrolled in Generative AI with Large Language Models from DeepLearning.AI. 

What were your initial thoughts about GenAI?

As someone with technical expertise in data science & ML, who also experimented with earlier generative models (e.g., GANs), I was quite excited and a bit surprised about how well LLM applications were performing. However, I had concerns about using GenAI, especially LLMs, reliably in real-life business applications. The volatility and limited explainability of these models, caused by their complexity and massive training data, worried me especially as I guide clients as a Senior AI Strategist at a German consultancy.

Which course did you take?

The “Generative AI with Large Language Models” course by DeepLearning.AI not only solidified my knowledge but also addressed my concerns while highlighting budget-friendly ways teams can apply GenAI.

How are you thinking about GenAI now? 

Surprisingly, the course gave me more hope regarding the democratization of GenAI. It demonstrated how small, open-source models could perform very well in certain use cases with adequate fine-tuning and alignment techniques. This approach gives developers and companies more control over their models, making GenAI accessible with a reasonable budget.

Are you still concerned?

The course deepened my knowledge in aligning LLMs with human values and introduced the latest evaluation metrics. While my concerns on trustworthiness remain, learning about these metrics was crucial. It’s an active field of research that needs more attention, especially considering upcoming AI regulations like the EU AI Act.

How does what you learned impact your work?

As a consultant and AI strategist, this knowledge allows me to better evaluate the suitability of various models and approaches for specific companies and use cases. It filled some knowledge gaps, enabling me to more confidently explain LLM concepts to colleagues and clients.

What advice do you have for others?

For those unsure about GenAI’s place in the workplace, this course offers a great mix of theory and practice. It provides the understanding needed to leverage GenAI effectively in business settings. So, if you’re looking to boost your knowledge in generative AI and large language models, my advice would be to go ahead and take this course ;).

Note: This interview was condensed and edited for readability.

Ready to take the next step with GenAI?