ClearPoint Co-Founder Bain Hollister and GM Microsoft Engineering Malen Hurbuns recently sat down to discuss AI and Large Language Models (LLMs) in a ClearPoint Unlock Podcast.
They looked at the growth and impact of AI globally, drew parallels to previous technological shifts and explored the adoption of AI and, in turn, the significant productivity gains for organisations and engineers.
A fast-moving change and its impact on organisations
Since the launch of Chat-GPT in late 2022, the technology industry has seen enormous amounts of investment and start-up activity in artificial intelligence and large language models, with new announcements every day.
The vast increase in this adoption and overall interest should not discount the many years of development behind AI and LLMs but comes with it now being seen as a very high-value part of the industry, providing ample opportunity.
“There are these big moments that happen in our technology industry, and they're pretty profound, and they have an impact on all of us,” says Hollister.
“How do we harness these things? Can we make ourselves more productive? What sort of impact will they have on our and our day-to-day work?” asks Hollister.
With it being so fast-moving, organisations and business leaders would be wise to stay informed on any developments to opportunities, risks or limitations to ensure they have an up-to-date understanding when making decisions about adoption and integration into their business and technology strategies.
“I think for a lot of leaders, it's quite a challenging conversation,” says Hollister. “It's easy to underestimate how difficult it is for leaders because there is a level of vulnerability in things that you don't know, that look quite game changer in their nature.
“The technology is here, but where this technology unpacks and where the experiences start to come through is still coming. It's a technology innovation that captures the imagination, it’s quite exciting but it does mean the conversations are quite wide-ranging, and there are people all over the spectrum about what they think may happen.”
While the conversations and approaches to AI sit across the spectrum, adopting these technologies is not without challenges, and organisations should ensure to adopt a responsible AI framework and policy before members of the organisation start exploring AI capabilities.
One of the main concerns for business leaders is the ethical implications of AI. As AI and LLMs become more advanced, there is a need to ensure that they are used responsibly and ethically – especially with issues such as algorithm bias, data privacy and inaccuracies.
Implementing a responsible AI framework provides guardrails for both an organisation and members of the organisation to safely navigate complex challenges and innovate responsibly, not only adhering to ethical obligations but providing a strategic imperative for sustainable growth and competitiveness.
Leaders looking to adopt AI-powered tools within their organisation and foster a culture around that should look to their mission, purpose and values to guide them on where the value lies for them.
“It’s a conversation that we have quite often with our colleagues and stakeholders in the industry – step one is learning,” says Hollister.
“We're lucky in our industry because learning has always been a competency – we've always had to stay current and learn and this one is, maybe it's quite a big development, but making sure that curiosity is here from a leadership perspective and that we're open and walking towards that is a key step.”
The changing face of software engineering
Where many leaders are still at the initial stages of understanding how best to utilise the many avenues that this technology will surface in and what this means for organisational productivity, software engineers have been early adopters and are already seeing big strides in performance and the developer experience.
“We have seen that software engineers are early adopters of such technologies with the adoption of AI-powered software development tools such as GitHub Copilot or Azure AI Studio in particular”, says Malen Hurbuns, GM of Microsoft Engineering at ClearPoint.
Software engineers can now provide increased value with AI tools automatically detecting bugs, suggesting fixes and automating everyday tasks, allowing them to work at a higher standard and focus on the real problems at hand with minimal context switching.
“When this wave of AI adoption started to boom, people anticipated job cuts for engineers because the machine could do all the coding, but what we are seeing instead is the sense that having some things automated away provides new opportunities to grow into new areas,” he says.
This can already be seen in prompt engineering, where instead of writing lengthy code scripts, developers can now provide a detailed prompt of the desired functionality, and the language model can generate the code to fulfil that requirement.
“The LLM is on my right-hand side, and the prompt engineering is on my left. It's not like Google where you put in the question and you get an answer, to be able to leverage or maximise the large language model, we also need to scale up and increase the equation on the left-hand side,” says Hurbuns.
When you're dealing with something as intelligent as a large language model, having strong prompts and learning the techniques behind them is one of the best ways we can leverage the technology, but those utilising it must ensure they consider and address any risks or challenges it comes with.
While the productivity gains for software engineers are already extensive, the industry is still in the early stages of seeing how this will tangibly impact teams.
“Right now, we’re seeing engineers on the individual level benefiting from coding and productivity gains. However, there still needs to be more adoption on the squad basis, and that kind of thing comes from your companies as well – how they adopt AI so that everyone gains,” says Hurbuns.
“We've seen companies move towards Microservices containers as well, but now is the opportunity to maybe swap out one of those Microservices with an LLM. It can do the same thing, right? I think a combination of those will actually be what the future of architecture will look like.”
Explore our interactive ClearPoint Technology Radar to discover the latest trends and insights, alongside understanding where specific AI tools, frameworks and platforms sit on the radar.
Looking ahead: organisational adoption of AI and LLMs
At ClearPoint, teams approach this fast-moving technology with a focus on two areas – thinking about it in an abstract way that considers the big changes coming and in what dimensions that will create impact, alongside fostering a culture of tangible exploration and storytelling, within the guardrails of safety.
“Being native software engineers from the very beginning is quite helpful for us because we can understand the fundamentals of what sits behind AI and LLMs and think in quite practical terms, but also in strategic terms around how we can bring this into clients and to our own ways of working,” says Hollister.
Teams are now faced with needing to identify between what looks like a breathtaking demo but may also be hyper-oriented, to the more significant changes and developments occurring and how they can best position themselves for those.
“Organisations who have been through digital transformation or have agile ways of working are going to be on good platforms to really leverage large language models, whereas companies that are still running on VMs or on-prem may take longer to transition,” adds Hurbuns.
“I think a key step organisations need to take is that reflection about ‘where do you want to be?’ There's opportunity here, do you want to take advantage of that or not? And if so, how should you take advantage of it?”
As organisations navigate this fast-changing landscape, strategic thinking is essential. There are challenges and opportunities presented by AI and business and technology leaders should make informed decisions about its adoption and integration.
ClearPoint’s approach, which is rooted in a culture of exploration within guardrails, reflects a deep understanding of the transformative potential of AI and LLMs. As software engineers experience substantial gains in productivity and performance, we aim to remain at the forefront, assisting organisations in responsible AI adoption.
The journey ahead involves not just technological advancement but a strategic reflection on where organisations want to be and how to leverage the transformative power of AI.
Get in touch to understand how we can support you find practical ways to leverage AI and use LLM in your business, so you can make the most of the opportunities these new technologies offer and add real value to your business.