Article

AI Insights: An interview with Elin Hauge

September 5, 2024 8 Minute Read

By Jen Siebrits

AI Insights An Interview with Elin Hauge

In a recent interview, Jennet Siebrits, Head of UK Research at CBRE spoke with AI strategist Elin Hauge to discuss the fundamentals of AI and its potential impact on businesses. Explore Elin's insights on the future, challenges, and ethical considerations associated with AI in the first part of this two-part interview below.

What exactly is AI? 

“The technical explanation is that Artificial Intelligence is mathematical algorithms in the form of iterative computer programmes (aka machine learning) that grind through large amounts of training data to build stochastic models. These stochastic models are then applied to new data to automate decisions (prediction models) or generate outputs (generative models). For prediction models, the training data can be your internal data, such as data about customers, products, operations, or maintenance schedules. For generative models, the models are typically built by external providers (aka foundation models, such as GPT-4 from OpenAI or Llama 3.1 from Meta), and then you will need to tune them on your internal data.”

"The simple explanation is that AI is mathematics and computer programmes applied to large amounts of data, including text to interpret, predict, or generate outcomes."


What is the potential of AI? Where could it have the biggest positive impact?

“The sky is the limit, with the limit being us humans! I hope we will prioritise the use of AI in industries with significance for the life and health of our planet and the humans and animals living here. Healthcare, waste management, energy, food production, transportation, and education are such examples.”

What is the current biggest blocker for AI development?

“There are probably more opinions than experts on this topic. However, from my perspective, the biggest threat are boards and C-suites without the relevant competence about the AI toolbox and its opportunities, limitations, and weaknesses. Without this competence, strategic discussions are likely to miss key inflection points, risk assessments will be lacklustre, sustainability reporting will have significant white spaces, cybersecurity threats will be overlooked, and compliance requirements will remain unattended.” 

What are the risks of AI?

“As I have explained already, AI is mathematics applied to large amounts of data to interpret, predict, or generate outcomes. All of these outcomes come with a probability of being correct. By definition this also means that they also come with a probability of being incorrect. In some cases, being a little bit wrong might not matter. For example, if Spotify’s listening recommendations are slightly off now and then no one really cares. But if you are the head of a social services function and you use AI to decide who does and doesn’t get benefits, the consequences of a wrong decision may be severe.

"We need to remember that AI is trained on data representing the world as it was. AI can only predict based on the data it is fed. AI has very limited knowledge otherwise. In that way, outcomes can be skewed (sometimes on purpose), given the data AI is fed.

"We need to consider that AI will undoubtedly be used by some for less altruistic purposes. To help mitigate the negative consequences, we need to build awareness about how AI works, and not to take everything we see online at face value! Also, we can demand transparency and accountability from companies and public institutions around the use of AI and the data used for training. We need to shine the light on how AI is used in various contexts and demand transparency and accountability.”

African American woman using a tablet

Artificial Intelligence

The adoption of AI is increasing, and leveraging its capabilities presents many potential benefits for real estate. Delve into our series to understand AI in context and discover its practical implications for the sector.
Explore the Series

Contacts

Related Insights

Item 0 Item 1 Item 2 Item 3 Item 4 Item 5