In recent times, ChatGPT has emerged as a revolutionary productivity tool, transforming how we perceive AI. Whether aiding children’s education or streamlining office tasks, it has undeniably become an integral part of our lives, akin to the internet’s transformative effect 25 years ago. This powerful AI is a far cry from the sentient beings portrayed in science fiction, like “HAL,” “ORAC,” or “Agent Smith.” Instead, it functions as a sophisticated “neural language calculator,” generating output based on vast amounts of data it has been trained on.
While some fear the potential consequences of AI capable of original thought, known as the “singularity,” the current reality is that AI serves as a valuable productivity tool. Rather than being self-aware entities, present-day AIs excel at predicting and generating text, crafting coherent sentences, paragraphs, and even entire articles in the style of known authors.
The tech industry’s focus on AI has shifted towards greater openness. Meta’s recent release of the Large Language Model (LLM) Llama 2, branded as “open innovation,” has sparked significant discussions. However, it’s important to distinguish this from traditional open-source software. Llama 2 comes with an Acceptable Use Policy to ensure responsible usage and prevent harm. This kind of openness is vital for building trust and addressing potential regulatory concerns.
Despite this openness, not everyone is rushing to download Llama 2 and build their own LLM. The scarcity of individuals with AI skills remains a challenge. The number of people capable of building AI apps, automating AI systems, or understanding how ChatGPT works is significantly limited. Meta’s decision to open up Llama 2 could help address this shortage by allowing more technologists to access and develop their skills with this advanced AI technology.
Llama 2 enables AI developers to build upon it in an open and transparent ecosystem, akin to the collaborative nature of GitHub for coding. This visibility fosters trust and encourages innovative advancements. For countries like the UK, gaining access to LLM technology can be transformative. British Prime Minister Rishi Sunak’s vision of the UK as a world leader in AI development can be realized through greater access to advanced AI capabilities like Llama 2.
Historically, closed LLMs hindered the pace of innovation, limiting the potential for skilled individuals and smaller companies to contribute. Earlier this year, when the original Llama was leaked and shared with researchers, it resulted in groundbreaking advancements that outpaced individual companies’ efforts. Opening up LLMs eliminates such barriers and allows for a more diverse and dynamic AI ecosystem.
Some critics express concerns about the risks of open AI, comparing it to handing out templates to build harmful devices. However, the benefits of open access to technology have long been evident, demonstrated by the success of open-source software that underpins the internet, the cloud, blockchain, and, indeed, AI itself.
Although Meta hasn’t fully open-sourced Llama 2, granting access to it signifies a step towards a more inclusive digital future. To lead in the AI domain, countries like the UK must prioritize understanding the need for openness and move beyond skepticism. As Rishi Sunak approaches the autumn summit, embracing the power of open AI will be crucial for realizing the full potential of AI and ensuring a better digital future for all.
