Dismantling AI Myths & Hype

Request edit access

Join in the effort to debunk and counter the hype, myth and inaccuracies around 'artificial intelligence'

Why the hype?

Everyone is talking about aritificial intelligence, but our conversations are riddled with hype, myths and inaccuracies. Even the term AI is mired in disagreement and deception.

Many people and organisations are working to ensure that the various technologies and systems we refer to as 'AI' are developed in a manner that respects our human rights, promotes social justice, and empowers vulnerable people. To do that work, however, means constantly fighting against hype, myths, and inaccuracies about AI that blind us to what is really going on, or twist the narrative in favour of certain interests.

For my Mozilla Fellowship project, I will be working to challenge these narratives, debunk these myths, and counter the hype and inaccuracies with solid arguments and facts. As part of this process, I want to open up a consultation to the public to find out which myths cause the most harm, and what we can do to counter them.

And please, if you think one of the 'myths' listed is actually true, feel free to get in touch and defend it!

Please contribute to this task by answering the questions below, or feel free to get in touch here: https://daniel-leufer.com/contact/

Daniel Leufer, 2019-20 Mozilla Fellow hosted by Access Now
https://www.accessnow.org/profile/daniel-leufer/

What are the most damaging myths, inaccuracies or bits of hype about AI?

The vagueness of the term 'artificial intelligence'

Sci-fi speculation about Superintelligence / The Singularity

The straight line of development from narrow AI to 'artificial general intelligence'

Ascribing agency to 'AI systems' (e.g. 'AI has discovered a new drug')

Fourth Industrial Revolution & other narratives of historical inevitability

Visual misrepresentations / stupid or misplaced robot pictures (see below)

Humans are also black boxes, so we shouldn't worry about AI black boxes

Optimization makes everything better

More and better data will solve any problem

Regulation of AI will kill innovation

Bad AI metaphors (e.g. "AI is the new electricity")

AI ethics washing / Ethics will be sufficient to govern AI

The 'AI Arms Race' narrative

We hold AI to a higher standard than humans, and that's unfair

AI inevitability (e.g. "if we don't develop it, someone else will and we'll have to buy it")

Politicians don't know anything about AI so government can't regulate it

AI will create new, creative jobs v.s. the reality of MTurk-style gig work

AI can solve any problem ("if you only have a hammer, everything looks like a nail")

Neural networks work just like the human brain, so AI will be able to do anything a brain can

Other:

What type of resources would help you to combat these myths in your work?

A database of detailed refutations / counter examples

Accessible summaries & overviews of academic and scientific opinions on these topics

Snappy refutations and short explainers of counter-positions

Other:

Do you know of any already existing work that challenges these myths and inaccuracies? Please add links or contact details for people we should get in touch with, we'd love to collaborate and showcase existing work!

Are you interested in collaborating yourself? If so, please explain how we could work together.

Please include your email address if you'd like to, or get in touch with me directly: https://daniel-leufer.com/contact/

Submit

Never submit passwords through Google Forms.

This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy

Forms