Latest AI Trends: Large Context Windows, Hyper-Personalization
An Interview With Maciej Micha?ek, AI Advisor and Principal Consultant for the Data Science & AI Practice of Lingaro Group
Lingaro:?Maciej, you’ve helped build AI solutions for common use cases for multinational companies and have helped business leaders find use cases for GenAI. As someone who’s an active participant on the field, what do you see as the latest trends when it comes to developing GenAI at the moment?
Maciej Micha?ek (MM):?One of the latest developments is that transcripts have now become more viable as source material for large language models (LLMs). This became possible because developers have expanded the context window for generative AI (GenAI) models. The latest model for Gemini, for example, has a context window so large that you are able to fit two volumes of Harry Potter there!
LLMs today have a good cognitive ability to understand the language and the intention of the user's query. From large collection of texts like meeting transcripts, the models are able to extract exactly the information you need.
Lingaro: The past trend has been to use LLMs to power search engines for internal knowledge libraries and help generate summaries and reports for managers. What does being able to parse hours-long transcripts mean for businesses now?
MM: In the corporate world, the new large multimodal models (LMMs) create completely new possibilities for information management. Every organization somewhere accumulates knowledge, so if a large company doesn’t utilize LMMs yet, they should start catching up as soon as possible.
GANTT by Lingaro version 3.0: What’s New
GANTT by Lingaro is Power BI Gantt chart extension that enables project managers and their teams to use a wide, comprehensive view to track and visualize multiple projects.?
GANTT by Lingaro enables users to customize their Gantt charts to suit their needs. If you’re new to custom visuals, read Microsoft’s guide on how to import custom visuals into the Power BI desktop.
Video: Introduction to Working With Power BI Gantt Charts
领英推荐
Well-loved features
Past features that carry over to Version 3.0 include the following:
Intersecting AI and Sustainability
Sustainability covers a broad range of interconnected issues ranging from deforestation to farmers’ livelihoods to making a profit. To tackle the challenges posed by sustainability, technology companies have developed tools that are sophisticated enough to help us become more aware of our impact on the planet and the people affected by our business practices. To help us process all of the data those tools make and turn that data into actionable insights, we use AI.?
Gen AI is particularly useful, given how it can learn from vast amounts of data and become able to create projections of the future. For example, imagine having to manage a fleet of trucks and determining their best routes in terms of fuel efficiency and customer satisfaction. Given high-quality training, a Gen AI model optimize supply chains, maximize resource utilization, and enhance particular functions that propel transformative change.?
However, training and implementing Gen AI requires a lot of energy. According to a study by researchers at the University of Amherst in Massachusetts, training a single model could produce over 280 metric tons of CO2. That was way back in 2019 — and it didn’t cover the carbon footprint of using the model. Given how AI models are much more powerful and popular now than back then, we can safely presume that AI is emitting way more carbon dioxide than it did before. Side note: We don’t have new figures because tech companies don’t disclose the carbon footprint of their AI solutions and there are currently no standardized methods for attributing emissions to AI training and utilization.?
Other environmental issues, namely the consumption of large quantities of water to cool down servers, the impact of extracting the precious metals needed for hardware manufacturing, and the disposal of burgeoning hardware waste must also be addressed.?
Got insights to share? Let us know in the comments.