GenAI: The New King of Productivity or the Emperor’s New Clothes?

Is it time to jump on the AI bandwagon, and who’s steering it at your organisation?

The last twelve months have seen excitement around Generative AI (GenAI) break out from the fevered imagination of internet forums to the front pages of newspapers. GenAI is going to revolutionise business, enable an unstoppable torrent of computer-generated cat memes along the way, and quite possibly make you redundant—even the CEOs! Almost half of CEOs think that their jobs could be mostly automated by AI ^[1]. Despite the hype, the sudden emergence of a new, epoch-defining technology feels wearingly familiar. Should we really care? Isn’t this just the latest example of a zeitgeist, willed into existence by some overzealous investors and tech companies, trying to find the next big thing amid declining revenue streams elsewhere?

Well, perhaps one useful way to understand what GenAI is, and is not, is to compare it to other “hyped” technologies: those that are now almost universally adopted (the cloud, big data), and those which are … not (blockchain). It’s also perhaps useful to use the implementation of previous tools as a guide to inform the steps organisations should be taking right now around the deployment of AI, lest it becomes a niche tool buried away in the basement or, even worse, a liability.

I like it! What is it?

A first similarity that GenAI shares with the technology booms of recent years is that, at least initially, nobody really understood what the tech was. In the case of blockchain, they probably still don’t. The unique aspect of GenAI is the ‘generative’ component — the ability to create new content based on simple user prompts. That could be creating an image for a document (like the one at the top of this page, created using Midjourney), drafting the executive summary of a report, or potentially even composing this article! Despite a lot of clever processing behind the scenes, which can make it seem like a ‘black box’, the key element in each case is that a human has guided the GenAI tool in terms of the output produced, though the AI has performed most of the heavy lifting.

The GenAI boom seems to have caught many people off guard. This is nothing new to either business or government sectors. It took a while for regulators to catch up with the ‘Big Data’ and ‘Cloud’. While we are now accustomed to mindlessly batting away pop-ups asking about cookies, the legislation that introduced this joyful concept (GDPR) was only enforced in 2018, years after widespread adoption of ‘Big Data’ practices. Likewise, Cryptocurrency, one implementation of Blockchain technology, is still discussed by government as if it were alien, despite Bitcoin having been around since 2009.

GenAI is in a similar early phase, where businesses seeking to best understand how to implement new technology safely and legally will have to accept that the ‘rules’ are in flux and likely to change frequently. However, the last thing any organisation wants is to introduce further ‘tech debt’ or deploy lawsuit-attracting technology for minimal benefit. The big data era had this same dilemma. For successful adoption, both technologies require organisations to carefully think through roll-out strategies covering leadership readiness, ethics, governance, communications, and people management. Who’s thinking about these issues with relation to GenAI at your organisation? What emerging policies or risk strategies can you point to?

“For successful adoption, both technologies require organisations to carefully think through roll-out strategies covering leadership readiness, ethics, governance, communications, and people management.”

Trust, but verify.

Another reason for such careful strategizing stems from a further similarity between GenAI and Big Data: public perceptions and the need for transparency. From the early days of Big Data, there has been public unease about the idea that governments and corporations have been gathering vast amounts of data and using it in ways people didn’t fully understand or agree with. Likewise, polling suggests most people don’t understand what AI is, don’t think that tech companies will develop AI responsibly, and doubt about effective government regulation of AI ^[2]. Notably, twice as many people are pessimistic about the future of AI as are positive.

However, the use of bulk personal data has been equally controversial, particularly for areas such as healthcare, but polling suggests when people are given a broader explanation for how their data will be processed and used, the vast majority can be supportive ^[3]. It will be equally important to educate users and customers around the beneficial impact of GenAI, and this is likely to become a focal point in the regulation of AI. What is your organisation doing to reassure users and customers around your use of AI, and to ensure transparency and openness?

Another consequence of the uncertainty around GenAI, and some of the reluctance to deploy it at enterprise scale, is the security of GenAI platforms and tooling. These same themes played out in the migration from physical data centres to cloud-based providers: “Where is my data?”, “How do I know it’s secure?”, “How do I know that it won’t be transferred to third parties without my consent?”. While such concerns have largely now been mitigated with respect to cloud providers, emerging best practice for GenAI, such as using services like Microsoft’s Azure AI, are beginning to provide similar reassurances. Do you know who is using AI within your organisation, and whether they are using an implementation of GenAI via a trusted service?

To the moon?

We can’t yet know how the GenAI revolution will pan out, but ChatGPT’s rapid adoption in early 2023 was heralded as ‘the fastest growing consumer application in history’ with 100 million users in two months ^[4]. Over half of organisations are attempting to use GenAI in some form ^[5]. This meteoric growth is on a different scale to the other breakout technologies mentioned above.

The human impact should not be underestimated and has the potential to be vastly different from the aforementioned technologies. Internally, there will be a genuine step-change in terms of organisational capability, meaning that staff members are no longer required to perform certain tasks. Externally, customers may suddenly find they are no longer dealing with a person but a chatbot or automated process. Particularly in domains where people expect a certain level of human interaction, organisations need to be aware of the messaging around the use of GenAI, with clear procedures and policies in place to explain how decisions are made, how they can be audited, and redressed if necessary. Currently, the majority of organisations claim to have no explicit policies in place governing the use of GenAI ^[6]. Does your organisation have an AI policy, and ways to justify and explain your AI-driven decision-making?

A final difference with GenAI is the crossover from consumer to business use. GenAI feels like it’s on the same kind of trajectory as smartphones, whereby almost everyone will end up using the tech, for both personal and work use, without even thinking about it. This adoption curve means that we are headed towards a future where people will expect GenAI to be utilised, and potentially frustrated with organisations that don’t use it. The rollout of tools such as Microsoft’s ‘Copilot’ means that AI-driven workflows will end up being the default option in many businesses, whether or not they intend it to.

With great power comes great responsibility.

In summary, while there are some parallels with prior tech booms, GenAI distinguishes itself in the rate of adoption and human impact. With many organisations already deploying GenAI, or considering deploying, early signs suggest the potentially transformational effects, and that future iterations likely to be even more impactful. Most organisations will want to capitalise on the benefits, and at the very least explore the potential to ensure they are not left behind. Every organisation has questions to answer around how this roll-out is managed.

It’s equally clear that there are very real concerns among consumers and customers, and that government regulation will have to develop further. In tandem with adopting GenAI, businesses must carefully consider how and when to deploy this technology, paying close attention to governance and ethical considerations. It may be the case that only 20% of businesses currently have AI strategies and policies in place, but 100% of intelligent organisations will be looking to rectify that. If you’d like to know more about how Agilisys can help you get ready to deploy Generative AI safely and responsibly, contact us today.

References

[1] – https://press.edx.org/edx-survey-finds-nearly-half-49-of-ceos-believe-most-or-all-of-their-role-should-be-automated-or-replaced-by-ai

[2] – Yougov Survey, May 2023, accessed from https://d3nkl3psvxxpe9.cloudfront.net/documents/Internal_AI_230509.pdf

[3] – Yougov Survey, May 2019, accessed from

https://d3nkl3psvxxpe9.cloudfront.net/documents/NHS_patient_data_public_opinion_poll.pdf

[4] – Reuters, https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/

[5] – Gartner, https://www.gartner.com/en/newsroom/press-releases/2023-10-03-gartner-poll-finds-55-percent-of-organizations-are-in-piloting-or-production-mode-with-generative-ai

[6] – https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai-in-2023-generative-ais-breakout-year