The superpowers of AI come with great responsibility
JULY 10, 2023 • BLOG POST • ANTONIO NERI, PRESIDENT & CEO OF HEWLETT PACKARD ENTERPRISE
IN THIS ARTICLE
- AI will create superpowers for those who tap into it, fueling tremendous advancements for business and society, but with these great superpowers comes great responsibility
- Business and government leaders must work together to reap the benefits of AI without compromising data privacy, sustainability, or responsibility
At HPE, we believe AI must be sustainable, responsible, and enable data privacy
Artificial Intelligence (AI) is clearly becoming an integral component of our future, including its ability to turn questions into discovery, insights into action, and our unlimited imagination into reality. AI will create superpowers for those who tap into it, fueling tremendous advancements for business and unlocking the solutions to some of the most pressing societal challenges we face, including climate change, the energy transition, and disease.
But with these great superpowers comes great responsibility. In the race to apply AI, organizations run the risk of relinquishing control of their data and privacy, potentially compromising their long-term competitive advantage. Given the amount of energy required to power AI, those same enterprises may also increase their carbon footprint. And without an intentional strategy around the responsible use of AI, businesses may not apply AI ethically or in line with their values.
So, how can businesses and society reap AI’s benefits while avoiding the repercussions that may come with it?
AI Can Unlock Innovation to Solve Societal Challenges
AI will increase the velocity of innovation across every industry. It has already begun to usher in a new scale of productivity and enable us to break new boundaries, and its abilities are nearly limitless.
For example, when the COVID-19 pandemic hit in 2020, scientists immediately turned to supercomputers to accelerate research using AI to better understand and combat the new virus. The scientists created a detailed 3D model of the spike protein to understand the strain and test it for potential drug interactions. Leveraging AI, the team accelerated calculations from 24 hours to just eight minutes, allowing them to run 1.2 million experiments per day. This saved thousands of hours, millions of dollars and countless lives.
AI can also be used to assist a wind farm designer to optimize their models to harness more power for homes and buildings. It can transform the role of first responders by enabling them to predict and mitigate the impact of catastrophic weather events through sophisticated digital twin modeling. It can even revolutionize weather modeling on climate change, accelerate the quest for nuclear fusion, and more.
However, business and government leaders must work together to reap the benefits of these innovations without compromising data privacy, sustainability, or responsibility.
Data Privacy Becomes Even More Critical
Data is the new currency that powers the digital economy. The key to success is extracting value from it. AI is an ideal tool to accomplish this, but organizations must pursue AI in a way that leverages their data privately to avoid diminishing its value. Customers should choose an AI cloud provider that does not use customer data to train its own models or use the data as a product to be resold or utilized with other offerings.
Through HPE’s AI cloud, HPE GreenLake for Large Language Models, we enable users privately train, tune and deploy their own models, utilize pre-trained models of partners, or establish domain-specific LLM capabilities using their own data without trepidation about leaking intellectual property or other sensitive data.
Managing the Environmental Impact of AI
As companies pursue net-zero targets, CIOs face increasing pressure to meet sustainability goals that consider and address the increasing desire to use power-hungry AI systems. From training to tuning to inference, the amount of energy required to use AI is enormous and will continue to grow.
As an example, researchers from the University of Massachusetts, Amherst, recently investigated the model training process for natural-language processing and learned the process could emit more than 626,000 pounds of carbon dioxide equivalent. That’s equal to nearly five times the lifetime emissions of an average American car.
The rapid uptake of AI requires us to be more intentional about sustainable IT innovation. While AI holds the promise of helping solve climate change, we must avoid it contributing to that problem. We must address energy consumption at every level – and organizations must embed it into their AI strategies from the start.
That means using more sustainable data center and colocation designs that leverage renewable energy sources, such as green hydrogen or solar power and wind power – as well as liquid cooling of the computing systems. The warm water created by the liquid cooling waste can be used for growing produce in greenhouses or heating homes – which can make the footprint carbon negative.
It also means more efficient AI and infrastructure innovation to optimize the use of resources – the faster the supercomputer, the more efficient use of time and energy is deployed to run AI models. HPE has great experience in this space; HPE Cray supercomputers represent 6 of the top 10 supercomputers on the GREEN500 list, which ranks the 500 most energy efficient supercomputers in the world.
A Commitment to Using AI Responsibly
While AI has the potential to solve some of society’s biggest, most pressing challenges, we must harness and govern the power of AI in an ethical and responsible manner. HPE believes AI must be:
- Privacy-enabled and secure – respect individual privacy, be secure, and minimize the risk of errors and unintended, malicious use.
- Human-focused – respect human rights and be designed with mechanisms and safeguards, such as human oversight to prevent misuse.
- Inclusive – minimize harmful bias, ensure fair and equal treatment and access for individuals.
- Responsible – be designed to enable responsible and accountable use, allow an understanding of the AI, and enable outcomes to be challenged.
- Robust – be engineered to build in quality-testing, include safeguards to maintain functionality, and minimize misuse and impact of failure.
Every enterprise that seeks to use AI should establish an internal oversight group that understands and can help create guiding principles, evaluates use cases and technologies, and has the technologies in place to make assessments on ethical issues.
Designing a Brighter Future with AI
AI will be the most disruptive technology of our lifetime, and 2023 will be remembered as a historic technology inflection point. We must lean into AI to fuel the societal advancements it will make possible – while remaining vigilant in mitigating its risks.
It is incredible how much the world has transformed in recent years, and I maintain my optimism and confidence in a bright future accelerated by the superpowers of AI.
Hewlett Packard Enterprise unveils AI cloud for large language models
HPE: Your strategic partner for AI
Celebrating one year of achieving exascale with Frontier, world’s fastest supercomputer