The US May Require a Humorous Tug of Open Source AI to Outwit China
“The US Needs an Open Source AI Intervention to Beat China”
“It’s a disturbing vision of AI’s future, one in which open collaboration—which has fueled the swift rise of artificial intelligence since 2012—becomes another casualty of a cold tech war. But it may already be happening.”
It’s terribly unsettling, isn’t it? In one fell swoop, open collaboration, the white knight in shining armor that has sped AI towards a brilliant tomorrow since 2012, might just topple over. Quite melodramatic. However, it might not be solely a hair-raising hypothetical. Some ominous signs whisper it’s already in motion.
OpenAI’s decision to restrict access to its latest natural language model, “GPT-3”, has ruffled more than a few feathers. Once a champion of accessible AI, OpenAI’s sudden change in tune is as perplexing as attempting to understand James Joyce’s ‘Ulysses’ after one too many ciders. The organization has justified its decision with ‘concerns about misuse’, leveraging the veritable boogeyman argument: the threat of malevolent use of AI. It does make one wonder whether little green men are right around the corner.
However, this consequential decision hasn’t gone unnoticed by other influential players within the AI arena. China’s State-backed researchers, with their affinity for red, are developing an open-source AI project of their own called “Wudao”. As catchy as the term “Wudao” sounds, it’s not a new dance trend but represents a looming presence extending China’s foothold in AI dominance.
Acheng Zhang from Horizon Robotics says the model could “create a community that will help to make the scientific progress more rapid and impactful”. Now that’s a party everyone would want an invite to! Zhang’s assertion should hold profound resonance within the tech world, but with increasing protective policies, there’s a likelihood this open festive spirit might just strike midnight before everyone has shared a dance.
In grappling with the potential impact, let’s not forget Elon Musk’s ominously cautionary words about AI becoming an “existential risk to humanity”. For OpenAI, the ‘fear factor’ provides a convenient curtain to safely hide behind. But, it’s crucial to remember that AI should serve as a helpful tool for all, and not merely become a weapon for the privileged few.
As it stands, restricting access to AI advancements may seem like a protective measure but, my dear Watson, it could inadvertently lead to a fracturing of collaborative innovation and stifle the shared growth that AI has seen over the past decade. It’s akin to making exquisite, bubbly champagne available only to the connoisseurs, while the rest of the world thirsts for a sip.
So remember, this isn’t just about who gets to play with the biggest, most sophisticated AI robot. It’s about a shared future where everyone gets the chance to solve Magic Eye puzzles, rather than accepting doomsday predictions about AI. That, perhaps, is the most human aspect of artificial intelligence. After all, it’s not merely what AI is capable of, but who gets to decide what AI is used—or misused—for. Isn’t that a thought worth contemplating over a glass of bubbly?