Part 4 in a series of 5
In this five-part series, Firehead takes a look Frontier AI and the impact it is having on technical communicators’ roles and the future labour market for digital communication
Trustworthy AI needs Trustworthy Digital Communicators and Content Creators
“As the field evolves, we must remain dynamic, learning to collaborate with AI rather than seeing it as competition,” says the UK government. Their stance on frontier AI is cautious overall, acknowledging its darker side and recognising the unpredictability of the technology:
‘Unchecked AI holds the power to assist in cyber-attacks, contribute to the
Gov.UK
creation of destructive weaponry, and propagate harmful content.’
A similarly apocalyptic message emanating from an international law firm warns: ‘While frontier AI promises innovation and efficiency, its dangerous capabilities are alarming…and if misdirected, have far-reaching consequences for public safety and global stability.’ Read on…. Much of the fear associated with AI is due to the unpredictability it wields; despite all the tech-savvy surrounding the models and algorithms, the outcome remains unquantifiable.
New AI can surprise us, showing dangerous traits despite thorough testing. Yet the UK GOV strategy has been to avoid strict legislation, instead encouraging a self-regulatory model, and within their paper titled Future Risks of Frontier AI they do cede a section to ‘Trustworthy AI’, with a concise definition: ‘Trustworthy AI is a combination of technical and non-technical approaches that cover all stages of an AI model’s development, from data preparation through to deployment.’
Part of the definition demands that the inner workings be transparent and allow human intervention at all times. Plus, the AI need to undergo test performances by developers to reduce the possibility of harmful output. This will require highly skilled content experts and technical communicators who can bridge the language gap between algorithms and the management, who must report on progress to the board.
In 2023, there have been clearly identifiable improvements in the capability of generative AI, including those identified by the UK Gov report:
- More complex, structured and accurate text
- Higher quality images
- Improvements in video, audio and 3D creations
- Fewer ‘hallucinations’ from AI fishing for red herrings
The UK Gov report continues: ‘Similarly, novel architectures, training, and fine-tuning approaches are being explored to enhance planning and reasoning capabilities… An initial step towards self-improvement could be the use of AI to produce training datasets or provide feedback to models during reinforcement learning.’
Superficially, this paragraph is incredibly positive for the technology workforce at the forefront of frontier AI. But it also sounds inevitable that we will soon be relying upon AI to police AI.
Firehead is incredibly optimistic about the future of digital and technical communicators and offers individuals and corporates the opportunity to upskill NOW to take advantage of these quantum shifts in the digital revolutions.
Contacts us here today to discuss your training requirements.
