Few websites are as essential to the internet as Wikipedia.

This digital encyclopedia was established by Jimmy Wales roughly 25 years ago while he was involved with the expert-driven initiative “Nupedia.”

Although Nupedia didn’t gain traction, its successor evolved into one of the most significant platforms online. Currently, there are over 7 million articles available in English alone.

So, where does Wikipedia stand as distinguishing fact from fiction becomes increasingly challenging and artificial intelligence (AI) transforms how we generate and access knowledge?

To explore this, BBC Science Focus held a discussion with Wales. He shared insights on the evolution of platforms over the last two decades, the challenge of sustaining trust and neutrality amidst misinformation, and how AI might shape the future.

SF: You’ve mentioned that social media contributes to a decline in trust. Can Wikipedia still thrive in such an environment?

JW: I believe we can certainly hold our own. I don’t necessarily view it as a competition, akin to business rivalries or transportation competition.

There’s a growing sentiment advocating for a shift from cultural conflicts to more thoughtful and respectful interactions, embracing differences, and promoting dialogue across divides.

Yet, division is an inherent aspect of human nature. For instance, when people discuss today’s youth, they often claim they lack attention spans and are perpetually glued to their phones. However, it’s evident that while they enjoy captivating short-form content, they can also binge-watch intricate TV series for extended hours.

Both represent facets of the human experience and can exist simultaneously—light-hearted social media and serious inquiry. Much like our dietary habits, we should ensure a balance, avoiding just junk food while incorporating nutritious options.

SF: How has Wikipedia managed to foster a relatively constructive culture while the rest of the internet seems to decline?

JW: I appreciate your use of “relatively constructive,” which recognizes that we face internal disputes and occasional anger—it’s part of being human.

It’s fundamentally about design—both in terms of software and what I term community design. It’s a blend of various factors.

When a business model relies heavily on ad revenue and user engagement, it becomes tempting to promote content that elicits longer viewing times, anger, or arguments, rather than happiness. This approach can be unhealthy in the long term.

Conversely, Wikipedia operates without ads and is funded by public donations. Thus, our model positions us as a charity, which means we aim to minimize clicks and keep users engaged without overwhelming them.

This necessitates a divergent approach to site design and community guidelines.

Civilized discussion is crucial. Without it, Wikipedia could become rampant with misinformation. I believe numerous foundational elements contribute to our success.

SF: Recently, Wikipedia has encountered increased criticism—Elon Musk even referred to it as “Walkpedia.” How do you feel about that?

JW: That’s unfortunate. When Elon labels us “Walkpedia,” he’s mistaken and misinformed. That assertion is baseless.

For instance, regarding transgender topics, there’s a common “gotcha” question: “What is a woman?” Wikipedia’s entry for “female” defines it as “an adult female human being.” This is indisputable, isn’t it?

Furthermore, the entry also addresses more nuanced discussions surrounding gender in society, which add depth.

Claiming that we’ve turned into some sort of radical leftist organization is simply misguided. Of course, there are areas worthy of refinement. At times, you may examine an article and feel it’s unjust towards one party, and these instances require critical reflection and effort.

The solution lies in engaging more individuals. I would welcome more compassionate and thoughtful individuals who recognize bias in Wikipedia articles to realize that contributions stem from source material, rather than assuming it’s the work of some radical activist ready to block dissenters.

While certain perspectives may lack representation, we can incorporate them and work toward a more balanced viewpoint.

read more:

SF: As reliance on AI for information grows, could Wikipedia face obsolescence?

JW: I hesitate to say “never,” as the future of technology is unpredictable. However, at present, we don’t foresee a significant impact in that regard.

A Pew Research study indicates that Wikipedia appears in about 3% of the top 10 results in standard Google searches, whereas Google’s AI summaries link to Wikipedia approximately 6% of the time.

This dynamic means Wikipedia is frequently referenced in Google AI summaries, although fewer users are inclined to click those links. We’re not overly concerned about clicks, but it remains a trend to monitor.

Personally, I utilize AI extensively and have discovered that its strengths lie in scenarios that wouldn’t typically generate interest on Wikipedia.

Forty years ago, I would have assumed that the first AI able to write in English would be dull and purely factual, merely regurgitating information. Surprisingly, I find that AI excels in creative brainstorming. It’s impressively effective.

SF: Are you concerned that Wikipedia’s biases could influence the AI models trained on its extensive content?

JW: Absolutely. This is a crucial focus, particularly for those developing AI models.

From what I’ve observed, the largest data sources for most advanced language models (LLMs) are Wikipedia and Reddit. While both are valuable, they have inherent limitations, particularly since both are predominantly male-dominated.

We are striving for awareness, as Wikipedia prioritizes neutrality. Yet, if we’re not vigilant, we risk unintentionally reinforcing existing biases.

At the same time, LLMs can be trained and directed, making it feasible to envision an AI continuously navigating through Wikipedia, identifying biases, and flagging articles in need of revision.

If utilized appropriately, this process could highlight instances where, for example, articles about male Nobel Prize-winning economists mention their families, while articles about female winners do not. Such insights are intriguing and may go unnoticed.

This conversation has been condensed for brevity and clarity.

Jimmy Wales’ new book, The 7 Rules of Trust, is available for purchase now.

read more: