You’ve talked about seeing a huge increase in demand for AI and engineering leaders over the last 18 months. What are you actually seeing from employers on the ground?
I’ve witnessed a couple of different things. Firstly, we've had more approaches from founders looking for Founding Engineers for their AI-native companies. Being AI native their first, and subsequent hires on the engineering side, need to be extremely AI savvy. People who understand AI infrastructure, and can build harnessed LLMs for a product. Secondly, there is a huge increase in demand for any skills that are AI-related from the larger, more established companies.
Four or five years ago we were inundated with demand for machine learning engineers, data scientists and data leaders, with companies heavily investing in building out their more traditional AI capabilities. Those AI & ML leaders then became the first to lead generative AI initiatives in their respective companies, equipping them with hugely valuable experience of generative AI in production. These same individuals are now in high demand as more and more companies invest in building generative AI-powered products.
Apart from this, we’re seeing a surge in demand for Forward Deployed Engineers and talent with experience in AI risk and safety. We expect both to continue to be growing categories.
Which AI roles and skills are currently proving the hardest for employers to find?
It goes without saying, but the people involved in pre-training, i.e. those who are involved in the academic experimentation and creation of novel models. The labs working on the frontier of this technology have something of a monopoly on the best talent, so we have a tough job on our hands.
We’re also seeing a need for individuals who are able to competently build and ship AI native products.These are skills that have only existed for a few years, hence there are fewer people experienced in this work. It’s a completely new skill set, so the pool is already small. On top of that, companies understandably desire people who have strong emotional intelligence, add to the company culture and have a skillset beyond the core AI needs.
Has the rapid growth in AI created a genuine shortage of experienced talent, or are employers sometimes being too specific about the type of AI experience they think they need?
In my experience I would argue the former. Our clients are pretty forgiving of the detail or context. What they want to see, especially in those early stage hires, is production experience. Individuals who have built something that was shipped, that is in use, and has been proven to work in the wild. There is higher demand as more AI companies pop up. There are only a fixed number of people with the necessary production experience, and that isn’t increasing fast enough to meet the demands of an accelerating market. Some companies might be willing to take a punt on candidates without direct production experience, but certainly in the early stages of building and shipping a product, the stakes are high. Therefore it’s likely that it will take a little while for the market to catch up.
What is absolutely critical for hires is curiosity and passion. People who are genuinely excited about where the technology is going and exploring and building personal projects with AI in their own time.
We’re working with clients building AI products in every industry and certain industries have more broad appeal than others.
Kelly Gauthier, Amplified People Group
What are the biggest mistakes you see companies making when they try to recruit senior AI and engineering talent?
The biggest mistake is a fixation on ‘brands’. We come across clients who become too focused on hiring someone who had worked at one of the big labs, or at one of the top AI companies. Being overly fixated on a specific brand or company name means managers and founders miss out on great talent that is flying under the radar. It’s tough to draw people out of the high profile companies, and there are so many interesting and exciting early stage AI companies that are not high profile yet. If you become overly fixated on that, you make your life unnecessarily difficult.
The other mistake is not understanding or embracing the market trends for compensation and offering packages well below the standard. For example, some clients have been doing machine learning for decades, and have moved into generative AI, as they have a great foundation for it. All of a sudden, these companies are now competing for talent. The AI engineers they want are the same people getting offers from the best labs and most exciting tech businesses in the world, who can pay three times as much!
When hiring managers are open to promising candidates from up-and-coming start-ups and lesser-known brands, they unlock a talent pool that has less pursuit by the market and, crucially, is more affordable. So we always try to steer our clients in this direction.
With so many businesses now competing for people with AI skills, what actually makes an employer attractive to the best candidates? Is salary still the biggest factor, or are things like the quality of the work, leadership and access to technology becoming more important?
Salary is certainly a factor, not necessarily the most important one, but it's definitely in the top three.
I think the nature of the product itself is probably the biggest factor. Candidates want to know: what am I building and what is the end goal? Candidates typically have a strong idea of what they are looking for when it comes to industry, product, and pay.
Of course we can’t forget culture & values alignment. We’re working with clients building AI products in every industry and certain industries have more broad appeal than others.
AI is also changing recruitment itself. Are you seeing AI specialists approach the hiring process differently, particularly when it comes to their expectations around how employers use AI in recruitment?
Nowadays, we're all using AI to augment the recruitment funnel - clients, recruiters and candidates. Given that we are all expected to use AI in our day to day work, we do expect candidates to leverage AI in their job applications and interviews, but it should not be used to replace critical thinking. Companies want to see candidates lead with personal experience, knowledge and skill, then layer in AI as a force multiplier rather than crutch.
For recruiters and clients, AI supports in a number of ways, from research to sourcing. Ultimately though great recruitment is all about partnerships and relationships which is something that AI cannot replace.
There’s a lot of discussion about AI replacing jobs, but at the same time you’re seeing significant demand for new AI talent. How do you see AI changing the make-up of teams over the next few years?
AI is still so new. It is early days and everyone is experimenting and building. Companies absolutely need AI talent to ensure that they don't get left behind and I don’t see this stopping any time soon - from AI-native businesses to more traditional businesses who are using it to improve efficiency and output. There are very few people with real experience in this sector. As it becomes more established and the pool of talent grows, there will be more supply to meet the demand.
Investment in this type of talent won't go away as individuals and companies keep innovating and pushing boundaries. As for the wider job market and what jobs will be replaced - predictions vary so widely it’s impossible to say what the future holds.
Finally, if you were advising a company that knows it needs to build its AI capability but hasn’t started hiring yet, what would you tell them to do first?
Be really clear on what they are hiring for, why and what success looks like. If they don’t have AI experts in the business, bring in external support to advise on the whole process, from the job specification and compensation, to the interview process and selection. Otherwise it can be a painful process and easy to get wrong.
There are plenty of people in the market who are extremely well-versed in how to build AI capabilities and products, and who also have commercial and product knowledge, who want to help companies. A lot of the data and AI leaders I've worked with over the last decade are now taking on consulting and advisory roles for exactly this purpose: going into companies, helping that company figure out exactly what they can do with AI, and the best way to do it. This is why I say: don't fly blind.
The most important thing to me right now is how to build AI safely. Set up your governance and guardrails right from the beginning, so that you're building something that is sustainable and robust, which will help and not harm your business or customers. There are so many steps to make sure your data is in order and your systems are safe. Equip yourself with experienced partners or team members who can advise you about safety and governance to ensure you have a strong foundation to build on.
