By Michaela Gordoni
Wikipedia co-founder Jimmy Wales says AI worsens the world’s trust issues, which can create a lot of issues across the globe.
“There’s been a long-term decline in trust,” Wales said at the WSJ Leadership Institute CEO Summit on Wednesday. The decline is “not perfectly even, but broad,” he added, which “leads to all kinds of problems.”
Generative AI, which can make fake videos look real, has destroyed trust in video, The Wall Street Journal reported.
“People assume it is false,” Wales said.
DigiCert reported that in controlled studies, people correctly identify AI-generated images around 62% of the time.
Wales said AI tools are “useless” for writing articles and isn’t worried about Wikipedia’s future being affected by it.
AI also affects trust between teachers and students.
Related: Why You Shouldn’t Trust an AI Chatbot With You Mental Health
“Those of us who teach find ourselves fighting for the minds and souls of students, clinging sometimes desperately to academic integrity and relevance, increasingly suspicious of those students — sometimes to the point of paranoia,” said one LA English teacher, Larry Strauss.
Cheating has always been around, but it has never been this advanced and this easily accessible. Teachers like Strauss have resorted to in-class pen-and-paper assignments, trading one headache for another as their desks accumulate piles of paper essays once again.
“Increasingly I am finding myself consoling and advising students who have been accused by their online college teachers of using ChatGPT and other bots to write at least part of their paper for them,” Strauss said.
Some of the students are genuinely dismayed by the accusations.
The teachers are “often impervious to my students’ assertions of honesty and authenticity. I am sympathetic to these college professors — and embarrassed by this AI derangement syndrome from which we all seem to be suffering,” Strauss said.
Many consumers (70%) also do not trust companies to use AI responsibly with their data, and companies’ AI agents sometimes operate without enough guardrails.
“Agents can execute tasks, access systems, and make decisions, often without clear identity, accountability or governance. Even organizations that know agents are operating in their environments may lack full visibility into their actions or clear accountability when something goes wrong,” DigiCert said.
Meta recently paid the price for this after hackers targeted Instagram accounts by successfully tricking Meta’s AI support chatbot into granting access.
These trust issues are all well founded. Society will need to find ways to preserve trust and improve accountability as AI tools continue to grow rapidly.
Read Next: Microsoft Joins NCOSE’s Annual Dirty Dozen List as AI Tools Enable Sexual Exploitation
Questions or comments? Please write to us here.
Share:
- Content: +1
- Content: +1
- Content: +1
- Content: +1