Hourly across the world is an avalanche of new digital products, platforms, digital businesses and services released to the market – all built around Chat GPT (and its variations) and now Generative AI technologies. These are just a few examples.
- AI Recolor changes the colour of any object in a photo.
- Youtube Summariser by Spext helps users quickly skim videos with auto-generated summaries, takeaways, and chapters for any Youtube video.
- Strapi Cloud is a managed content platform built on top of Strapi, an open-source headless CMS.
- ZapHire screens, validates, and sources potential job candidates.
McKinsey, the global management consultant, said ChatGPT seems to have trouble counting or solving fundamental algebra problems—or overcoming the sexist and racist bias that lurks in the undercurrents of the internet and society more broadly. Generative AI outputs are carefully calibrated combinations of the data used to train the algorithms.
This newsletter goes deep into the risks and opportunities of Generative AI.
The industries most affected – is where content is logically complex – and the final output of the complex information is critical to success. It’sIt’s professional services and marketing.
Investments
PitchBook found that venture capitalists invested $1.7 billion into generative AI startups in the first quarter of 2023, with another $10.68 billion in deals announced during the same quarter.
The problems
- Untested technology – it is learning in real time,
- Unsettled copyright and IP questions – who owns the output?
- Data privacy legislation – the potential for a business to inadvertently publish others’ private information,
- The potential for reputational and legal risks in unintentionally publishing other people’s original content with biased, offensive, or factually incorrect and copyrighted information remains high.
What is generative AI?
Generative AI is a form of artificial intelligence in which algorithms automatically produce text, images, audio and video content. The technology is taught via massive amounts of data and works by predicting the next word or pixel to produce a creative output.
The problems
- The risk to jobs: Generative AI produces human-level work. It is impacting the jobs market.
- Risk of fake content: Generative AI creates human-quality artifacts at scale, including fake and misleading articles, essays, papers and videos. Misinformation – fake news – is not a new problem: today and tomorrow, it will be at never-before-seen levels.
- It is a significant risk but may be manageable. Fake content could be identifiable with legislation that requires (a) watermarking identifying AI content at the point of creation or (b) deploying AI countermeasures that find misleading and deceptive AI content after the fact.
The Big Problem
The production of fake content on either a local or global scale maximises influence and persuades people, voters, and buyers to form wrong conclusions – not based on facts.
Targeted influence campaigns are now heat-seeking hypersonic missiles zeroing in on individuals for optimal influence outcomes. In real-time, the campaigns are updated in micro-seconds to increase the persuasion of individual people.
Every detail of a video, written content, and speech can be customised in real-time by generative AI to maximise influence campaigns’ impact on people.
Real-time tracking of user engagement, it learns which tactics work best for you. It learns how to get your attention to continue with engagement and deliver persuasive content to change your behaviour – for good and evil.
Targeted conversational influence is a generative AI technique conveying influence through interactive conversation. Chatbots are an example today. Tomorrow’s conversational influence is the ability to optimise every word to achieve an outcome.
How it works
Chatbots, videos and written content will understand your feedback and reservations in real-time. It will adjust its persuasion pitches to persuade people to purchase or vote a particular way.
Knowing in advance at a personal level how people respond to
- Logical appeals,
- Emotional arguments,
- Savings,
- Time-pressure,
- Free add-ons,
Ensures that the right strings maximise influence campaigns that deliver the desired change.
The potential rise of propaganda, fake news and misinformation to levels 100x more than today, which seeks to persuade people with false information and extreme ideologies, is a risk to society.
Consider this – a conversational agent (advanced chatbot) convinces you that an unsafe medication will fix your health problems. The chatbot can cherry-pick evidence in ways that would overwhelm even the most knowledgeable person.
It’s an age-old argument – good or bad.
Manipulation or persuasion – industrial scale influence requires Parliaments to catch up and step up quickly.
Into the void enters the new world of AI, and directors, boards, and business leaders need to have eyes wide open for future legal threats committed today.
Regards
David
