Empowering the use of innovative technologies to achieve sustainability goals
- Businesses have sustainability goals in their roadmaps, and there is a lot of focus on sustainability on a global scale in general
- Large language models should be discussed in relation to the expanding data universe
The right analysis can help organizations improve their impact and transform the way they do business to become more relevant to their customers in a meaningful and credible manner, which is especially important when it comes to achieving their sustainability goals. Data is essential for a company to truly understand their current status.
To do that, finding out where the best sustainability opportunities are will help.
Given the volume of data generated by an organization, the first step in investing in sustainability is to analyze the different types of data that are accessible and to determine which data would be useful to monitor sustainability for the company. The systems needed to collect and aggregate the appropriate information will then be determined.
Tech Wire Asia had the opportunity to sit down and speak with Dr Eng Lim Goh, Senior Vice President & Chief Technology Officer, AI at Hewlett Packard Enterprise, and Asst Prof Feng Meng Ling, Assistant Director of Research (Healthcare), National University of Singapore, about how emerging technologies are accelerating data sustainability and assisting customers.
The state of data sustainability on a global scale
Before engaging with how emerging technologies are thriving, it’s important to note where sustainability lies in the current world. According to Dr Goh, there is a lot of focus on sustainability on a global scale in general, to the point where businesses, whether they are from Asia Pacific (APAC), Europe, or the US, now have their sustainability goals for the years 2030, 2040, or even 2050 in their roadmap.
Many companies have actually linked executive compensation to sustainability goals on the roadmap in an effort to demonstrate their commitment.
Talent recruitment is the other aspect. He referenced an Intergovernmental Panel on Climate Change (IPCC) report stating that the globe is expected to hit the 1.5ºC level over the next two decades and claimed that only the most extreme reductions in carbon emissions from this point forward could help prevent an environmental catastrophe.
Dr Goh observed that “people are starting to internalize it, especially the younger generations. Unlike 20 years ago, sustainability is a topic that is extensively discussed today. Since [they will be looking at people like us] and wondering how we are contributing to the sustainability plan for the future, recruiting talent is becoming increasingly crucial.”
Organizations give talents a sense of purpose by embracing and helping employees with sustainability initiatives. In other words, businesses that can create a better future for their employees are likely to enjoy the best futures for themselves.
He explained how Science Based Targets demonstrates that companies are taking sustainability seriously through the Science Based Targets initiative (SBTi) and have shown companies how much and how quickly businesses need to reduce their GHG emissions to prevent the worst impacts of climate change, leading them on a clear path towards decarbonization.
The importance of sustainability in the expanding data universe
Large language models should be discussed in relation to the expanding data universe since they have the potential to have an impact on a variety of business operations, including marketing, sales, finance, legal, human resources, and even manufacturing.
On the basis of very sizable text-based data sets, large language models are machine learning algorithms that can identify, forecast, and produce human languages. They consume enormous volumes of the available data, and this consumption requires a lot of energy.
“When this was internalized, the companies [whose initial priority was to have these models built] are now considering the sustainability of these models. Why? They consume a lot of energy in learning. The models are large, and when they respond to inquiries, they use a lot of energy,” said Dr Goh.
Therefore, businesses must find a way to construct large language models in a way that maximizes energy efficiency. “Don’t forget about energy efficiency. Build a model that is not only efficient at learning, but also at providing answers,” he advised.
Emerging technologies to help APAC businesses achieve their goals
First, large language models, for instance, influence how they communicate with customers. Even call centers are now being interacted with by bots. By combining with the large language models, bots can be even smarter.
“It’s very likely that some companies will use [large language models with bots] to establish a more personal connection with incoming callers rather than just providing blunt responses. I think it will also affect how you communicate with customers,” he continued.
Second, building a digital model of the infrastructure(s) and running it on a computer is essential for energy efficiency since it allows for visibility into “what ifs” and potential improvements to the infrastructure itself.
Dr Goh stated, “Singapore is creating a digital twin of the city and running simulations to see how the air moves through the city, different buildings, and then to see how we can alter things so that we can cool the city by one degree.”
He also gave an example where, over a 50-to-100-year period, as various building permissions are requested, where buildings are being torn down and replaced, fluid dynamics experts look at it to say, “You’re tearing down this building by asking for new buildings to be built taller, when perhaps you should just adjust the building slightly so that the air will flow better.”
The workplace is represented digitally by “digital twins.” They map a facility’s space and resources in order to align them with utilization goals. Since digital twins frequently come with their own AI engines, a lot of AI simulation is used by organizations to fulfil their objectives in terms of sustainability.
How HPE technologies are benefiting customers
HPE has a wide range of customers from different industries whereby they benefit from the use of their technologies. However, among all of those customers, one struck out as being in the healthcare industry, FathomX, a medtech start-up from the National University of Singapore (NUS).
HPE and FathomX inked an MoU back in the year 2020 in order to work together to improve the effectiveness and productivity of healthcare in the APAC region. What was inspiring about this collaboration was the ability for FathomX to make breast cancer diagnostics easier and more accurate through AI.
Interestingly though, there were obstacles that the company had to go through to fully capitalize the power of AI. According to Asst Prof Feng, diagnosing a mammography for breast cancer is an extremely difficult AI task by default. The false positive rate can be as high as 10% to 15%, and the false negative rate can also be 10% high, even for experienced radiologists.
Additionally, he said that it is a really difficult assignment for AI. Such a complex AI model needs to be trained using a lot of computer resources, both during training and even during deployment.
“However, when combined with the HPE edge computing device, they enable us to equip with powerful enough GPU processing power, on the inference level, so that we can easily demonstrate AI to our clients. We will use more advanced High Performance Computing GPU basis computational power units from HPE, which enable us to speed up deployments with hospitals, when we have actual deployments in the hospital,” he explained.
He also explained that collaborating with HPE is beneficial for both companies. The organization may benefit from HPE’s experience in creating enterprise solutions and assist HPE in growing their presence in the healthcare sector.
- The largest deal made by Cisco is heating up the AI cybersecurity race. Here’s why
- Chinese electric cars can now be controlled with smartphones
- After Bard Extensions, the Microsoft Copilot AI companion unveiled
- Speaking easy: is realtime translation ready?
- WhatsApp for Business targets Indian enterprises