The environmental threat of single-use plastics is something we all know. From documentaries on topics to the introduction of wooden cutlery to restaurants, Katado plastic has been exposed as a threat to wildlife and our planet. As individuals and business, we were trendy reusable water bottles and plastic straws that were less common.
But there is another hidden crisis facing businesses today: single-use data. It means your data is frug arrogant – it's about it right away.
Data difficulties
NetApp defines a single usage data as data created once by a company, such as emails, attachments, or even GIFs, but is stored forever. That data will never be used again and will continue to consume energy in storage, which will affect the environment. This will accelerate into the age of AI as companies rapidly increase the amount of data they create, consume and store.
However, we found that we don't need a lot of data to fuel our AI. Importance lies not in quantity, but in quality. Data must be clean, organized and connected to AI ready. For that to happen, businesses don't have to stock up on data and can adopt a modest initial mindset without limiting the possibilities of AI solutions.
In addition to this, there is no denying that the UK is enthusiastic about leading the creation and use of AI among other countries (a recent survey found that 34% of UK respondents are ranked as leaders in AI innovation over the next five years). However, there is not just one way to “do” AI.
With the environment in mind, working on both single-use data while implementing a modest strategy in AI means that businesses will not compromise AI ambitions while using less data. Recent NetApp data has actually discovered that a majority (85%) of UK IT leaders know that data management can help reduce carbon footprint. So why aren't they taking an alternative approach to AI that uses less data? Certainly, is it the key to reducing emissions and enabling businesses to balance their lofty AI ambitions while achieving green goals?
Hidden costs of data
Let's start with the underlying causes. AI is undoubtedly creating an unprecedented demand for data. Our study revealed that most IT leaders expect data estate to grow by 41% to accommodate these AI ambitions. It's a fair amount of data, energy, and ultimately money. What's less recognized here is the incredible existing sizes of many data estates today. It houses these already long-standing corporate data, many of which are already unused, unorganized and unnecessary.
Such AI growth strategies focus on collecting more data at any cost, regardless of their true long-term utility for tools and technologies, making it easy to show the increase in single-use data. Not only will businesses have to pay more cash to store unwanted data, they will need to increase their carbon footprint, but the success of the AI initiative itself could also be threatened.
It is now well known that data is fuel and AI is engines. The cleaner the fuel, the greater the impact it will have on the engine. Quality data provides quality insights. Because of the data, data can result in AI results that take more time to spend in search of practical results than actual success.
That's where a considered data management approach comes into play. This distinguishes between successful AI solutions and flat-falling solutions. A planned and strategic approach to data growth will organize the data that the data business stores and make it more effective to use.
By making the data contained in the language model more selective, businesses can have better control over their carbon emissions and limit the energy they spend on their most important resources. For example, in healthcare, separating the latest medical information and guidance from other information about the topic means a safer, more reliable and faster response to a patient's treatment.
Being simple is the future path
It's been several years since the classic slogan “decrease, reuse, recycling” stuck in my mind. It may be a bit old now, but it fully captures the approach needed for AI.
Frugal AI means adopting an intelligent approach to data focused on using only the most valuable information. If your company has a better understanding of your data, you can significantly reduce the storage of single-use data by labeling it, identifying it, which teams are responsible for its removal. Only then can a simple AI system be implemented, allowing businesses to adopt resource awareness and efficient approaches to both data consumption and AI use.
It is important to emphasize here, but it means that simple AI does not mean that there are fewer end results or lower technology impacts. This means that data is concentrated, small and equally impactful, leading to AI. Think of making a drink with extra concentrated squash. Frugal AI is an extra concentrated squash that places data efficiency, considerations and strategies at the heart of an organization's AI ambitions.
When checking that a single usage data is finished with unnecessary data creation, companies need to shift their mindset about how data is presented. If you leave one thing behind, think of the administrator as being considered not as a disturbance, but as a stepping stone to better AI.
Key takeout
- Single-use data is used as data once created by a company, such as emails and attachments, but is stored forever.
- Data must be clean, organized and connected to AI ready.
- Organizations already have years of enterprise data, many of which are unused, unorganized and unnecessary.
- Data is fuel, AI is engine. The cleaner the fuel, the greater the impact it will have on the engine. Quality data provides quality insights. Because of the data, data can result in AI results that take more time to spend in search of practical results than actual success.
- A planned and strategic approach to data growth will organize the data that the data business stores and make it more effective to use.
- Frugal AI means adopting an intelligent approach to data focused on using only the most valuable information.
Kirsty Biddiscombe is NetApp's EMEA AI lead.
read more
Why Knowledge is the ultimate weapon in the Information Age – Learn how to build a human knowledge-first approach to AI, allowing organizations to execute the best possible information
Why ISO 42001 sets standards for responsible AI governance – With the use of AI, the development of effective governance is paramount, with the increase in all-all areas. ISO 42001 is the latest standard to help businesses build trust that moves forward
Why AI needs kill switches – Just to be safe, artificial intelligence is on the rise, and organizations are under threat. Here's how to put a kill switch on AI when you need it
