WRITTEN BY OMAR NASIR

Omar Nasir is a Data Scientist working in the Production Creation- and Future Lighting teams at Helvar.

 

 

We live in a rapidly evolving world, where the pace at which changes happen tend to be both exciting and overwhelming at the same time. Familiarizing oneself with new innovative technology and being proficient to the point where it can be applied for both personal and professional betterment is a daunting task, especially when the risk of the technology becoming obsolete in a short period of time is quite high. Fortunately, this does not need to be intimidating; we must learn to lean on these very tools and maximize their potential to not only improve our productivity and well-being, but ultimately become a better version of ourselves.

I must confess it is easier said than done. Almost a decade ago, self-driving automotive technology was (and still is) an exciting research area. We studied the underlying technology at university. We mapped out business cases and debated how the technology can impact the future of smart cities. Fast-forward ten years, and while the effectiveness of the current solution is still debated, it is an indisputable fact that the innovations in the field have led to a significant leap in our understanding of computer vision. It is perhaps easy to fall into the 鈥榚nthusiasm trap鈥 i.e. to have an inner monologue asking the question 鈥渨hen I can start working on this technology?鈥. It is a feeling that naturally fosters around an evolving landscape, but it is equally crucial to keep yourself grounded and ask the question:

鈥淗ow can we use this technology to improve our customer鈥檚 experiences?鈥

Staying focused when tackling an ever-changing frontier is difficult, but necessary. We all take advantage of multiple learning avenues, for example, keeping up to date with latest research papers and conferences, reading technology newsletters, subscribing to e-learning platforms etc. As we develop a deeper understanding of business models and gain domain specific knowledge, we naturally become better at extracting value from the mountain of research and innovation. Filtering out noise from a plethora of signals has been instrumental as a guiding principle for personal growth.

In the service industry, research and innovation around the topics of AI and Machine Learning often directly lead to an improvement in product portfolio and user experience. However, we can also leverage the data around us to improve our quality of life and professional expertise. For example, if you are a smartwatch owner, you are probably familiar with the variety of insights generated by the device. They are certainly useful, but they do not often tell the full story. One way to maximize their benefit is to extract historical data and analyze physiological markers such as RHR (resting heart-rate). I often monitor the RHR values, and if I notice that it is constantly elevated, it might be time to consider taking a short break to rest and recharge. This is, of course, only relevant after controlling for variables such as physical activity and sleep schedules.

It is 2024, and LLMs are the most popular technology in the world right now. GPT bots are inescapable; whether you are employed in the smart buildings industry, healthcare or academia everyone wants to tinker around with LLMs. I have used various chatbots and LLM based tools, and I believe that as long we stick to the principle of 鈥淚mmediately verifiable output (IVO)鈥, we can maximally harness the potential of LLMs, not as a tool to replace our efforts but to augment and improve our productivity. IVO is a very simple principle: if the output from an LLM can be immediately verified that it is correct to an acceptable degree, it can be confidently used as-is or with slight modifications. This can greatly speed up productivity and eliminate redundant tasks. A pertinent example of an IVO adjacent LLM is a coding assistant that assists us in writing boilerplate and repetitive code. For creative tasks I almost always go back to the drawing board, however, for rudimentary data cleaning, data visualizations and structural code generation, coding assistants can reduce the required time and effort by orders of magnitude, ultimately giving us more freedom to work on more challenging and innovative problems.

To learn more about Helvar’s working culture follow our Life at Helvar Instagram account!