The Power of GPTs
The Power of GPTs
And how businesses will need to adapt to take advantage of them
The biggest announcement made at OpenAI’s DevDay was undoubtedly GPTs. You can find my write up on the event here.
When Apple launched the App Store back in July 2008 it started with 500 apps, and 15 years later there are around 1.6m apps available in the App Store. With the launch of GPTs there are currently 15 available to try, it’s going to be interesting to see if GPTs follow the same trajectory as apps!
What are GPTs?
GPTs are AI 'agents' that are tailored versions of ChatGPT for specific purposes. They are built by combining instructions, expanded knowledge, and actions, with the ability to be published for others to use. Users can program a GPT simply by having a conversation with it. This makes it easy to customize the agent's behavior to match the desired context or task, enabling a wide array of applications.
Some of the main features of GPTs are:
Built for specific tasks: GPTs are designed to follow developer-defined instructions to carry out specific actions based on user input.
Extended knowledge: A GPT can be updated to contain specific knowledge beyond the default model, such as additional content or specific task-related information.
Integration with other actions: GPTs can call predefined functions allowing them to provide more sophisticated and specific responses.
Customisation: Users can customise their GPTs, making them more useful and effective for specific tasks.
Publishing and Sharing: Users can publish their GPTs so that others can use them or keep them private. They can also be shared within an organisation when using ChatGPT Enterprise.
Giving agency to everyone
When Sam Altman announced GPTs he focused on some very compelling headlines:
GPTs will make it easier for users to accomplish all sorts of tasks.
Users can program GPTs with language just by talking to ChatGPT.
It’s easy to customise the behaviour of GPTs so that they do exactly what a user wants.
Building them is very accessible and it gives agency to everyone.
It’s this last point that makes GPTs so exciting - they are doing what generative AI does best, democratising access to incredibly powerful capabilities. GPTs are a second generation generative AI technology (try saying that fast 🤓) - they’re more powerful and more accessible to more people than the foundational models they’re built on. This is why GPTs are so important and why they are going to drive the next phase of OpenAI’s growth both in terms of users and revenue.
OpenAI also announced at their DevDay that they now have 100m active users on the platform every week as well as 2m developers and 92% of the Fortune 500 company using their platform. This is all from a standing start a year ago - that’s some impressive scaling! I believe, that if the stars align, GPTs will drive x5 user growth for ChatGPT over the next 12 months, allowing OpenAI to surpass X’s (Twitter) monthly user numbers, and cementing its place amongst the large digital platforms.
However, there is one big challenge that OpenAI will have to overcome in order to realise this growth and that’s enterprise adoption. There is so much potential for this 2nd generation generative AI technology in the work place, but there are a number of challenges that organisations will face in adopting GPTs that might prevent OpenAI from seeing this x5 growth.
What are the challenges facing organisations in adopting GPTs?
In the UK, approximately 48% of all jobs are in small and medium-sized enterprises (SMEs), 32% in large mutli-national organisations and the remaining 20% of jobs are in the public sector.
In both large multi-national organisations and the public sector (where approx 50% of jobs are), the traditional approach to IT typically involves a centralised structure with a focus on standardisation, consistency, and control of IT resources. These IT strategies are often conservative, prioritising system stability and security over rapid innovation, with a focus on maintaining legacy systems.
In SMEs, where the other 50% of the workforce is employed, the approach to IT is often more flexible, but is highly cost-sensitive compared to large multi-nationals.
To summarise, in one half of the working population you have IT teams that are focused on standardisation, consistency, and control, whereas in the other half you have a high degree of cost-sensitivity. Across all of this however is a landscape ripe for the transformative influence of generative AI. While some organisations focus on control and others on cost, both are seeking efficiency and innovation - fertile ground for GPTs to demonstrate their value.
Despite these organisational constraints, a trend of generative AI usage in the workplace is already emerging. Many professionals are independently exploring generative AI tools to enhance their productivity. This grassroots adoption highlights the latent potential of generative AI, which, when fully integrated with organisational systems and workflows, could unlock unprecedented levels of innovation and efficiency.
Organisations open to rethinking their technology strategies stand to gain immense value and competitive edge from generative AI. Those slower to adapt might find themselves challenged by more agile competitors and newcomers designed with generative AI at their core. In the near future, access to generative AI tools could even become a sought-after attribute in job candidates, akin to the demand for hybrid-working options today.
2024 is going to be a really interesting year for technology in organisations. The arrival of ‘Enterprise-ready’ generative AI tools marks a pivotal moment. How businesses embrace and leverage this technology could very well redefine their future success and influence the broader technological landscape.
“The future is already here, it’s just not evenly distributed.“
William Gibson