What’s the worst that AI can do to our industry (and can we prevent it)? – Unmade: media and marketing analysis

Welcome to an end of week update from Unmade. Today, we share some proposals formulated at our HumAIn conference on how the marketing and media industry should be addressing the arrival of generative AI and the approach of artificial general intelligence. And further down, a better day on the Unmade Index.

If youve been thinking about upgrading to an Unmade membership, this is the perfect time. Your membership includes:

Member-only pricing for REmade (October 1);

A complimentary invitation to Unmades Compass event (November);

Member-only content and our paywalled archives;

Your own copy of Media Unmade

Cat McGinn, curator of HumAIn, writes:

Back in 2007 or so, an American high school teacher made a Youtube video in which he cheerily asks, on the subject of climate change action, whats the worst that can happen?

He goes on to sketch the Climate Change Decision Matrix, which assesses the potential outcomes of action or inaction on climate change, considering both the risks and benefits.

Were at a similar inflection point with AI.

Its been almost a month since our AI conference for media and marketing, HumAIn. During the conference, attendees collaborated on a set of principles for the industry, guidelines specifically designed to offer media and marketing professionals a path for responsible AI adoption.

In the weeks since the conference, high-profile staff responsible for safety at the AI company OpenAI have quit. Several senior AI researchers claimAGI - artificial general intelligence will be achieved by 2027.

Google Deepmind and former and current OpenAI staff have penned a joint call for protection for whistleblowers concerned about the speed and direction of AI developments. The Media, Entertainment & Arts Alliance called for legislation on AI as a matter of urgency, and even the Pope lectured world leaders at the G7 Summit on the risks of AI.

Theres a simple reason genAI offers such utility to marketing tasks: large swathes of the internet data these large language models were trained on were created for the purposes of advertising.

The humAIn AI Principles were developed using data from our survey of Australian and media and marketing professionals.

One of the strongest data points was the fact that almost 80% of respondents felt there was inadequate regulation around genAI.

In the continued absence of any robust response from our elected officials, self-regulation seems to me an act of enlightened self-interest.

On the day, our working group developed the HumAIn AI Principles as a starting point for the conversation on how we as the communications industry might engage with the development of AI. The Humain principles are":

Unmask Everything: Be brutally transparent in all AI applications.

Put Humans First: Design AI to serve people, not the other way around.

Reward the Good Guys: Financially back companies that stick to their ethical guns.

Teach the Masses: Make AI knowledge accessible to everyone, everywhere.

Trust, But Verify: Scrutinise every AI output and data source.

Spill the Beans: Share all successes and failures openly to help us all improve.

Play by the Rules: Follow all privacy laws and regulations to the letter.

Give Credit Where It's Due: Acknowledge everyone who contributes to AI development.

School the Team: Continuously train teams to be AI-savvy.

Champion Ethical AI: Promote balanced, ethical use of AI tools and tech.

Ditch Our Principles: Stick to our ethical guns, no matter what.

Chase Hype: Stay grounded and avoid the AI hype-train.

Worship Machines: Never let AI overshadow human importance.

Let Bias Slide: Actively hunt down and eliminate AI bias.

Use Dirty Data: Only use clean, legally trained AI models.

Replace Humans: Use AI to boost human roles, not replace them.

Compromise Privacy: Guard people's privacy and security fiercely.

Rip Off Ideas: Respect and protect others' intellectual property.

Create Noise: Avoid pumping out useless or misleading AI content.

Wait for the Law: Go beyond the law to ensure our AI ethics are top-notch.

There is a more detailed breakdown of the research and development process from Time Under Tension,including using AI as a research partner.

Based on the principles, I asked GPT4o to write a decision matrix for how we as an industry should be acting upon the arrival of genAI.

Heywhats the worst that could happen?

What’s the worst that AI can do to our industry (and can we prevent it)? – Unmade: media and marketing analysis

Categories

Archives