“I work in AI, and it is a nightmare”
I am a data scientist specializing in AI. I have a PhD with an AI focus. I have spent the past five years working in data science and AI. During that time I have had six jobs, none lasting more than 18 months.
I suspect that I am not alone. People think that AI is the place to work because everyone is hyped about it. What they do not understand is that the hype itself makes working in the sector a nightmare.
Part of the problem is that there are different definitions of AI and only some are currently sexy. The excitement surrounds large language models (LLMs) to the detriment of other equally valuable machine learning methodologies. My PhD didn’t cover LLMs, so I’ve had to spend some time upskilling myself.
The hype around LLMs means there’s no shortage of money available to fund LLM projects, but many of these projects are misconceived. A lot of companies want to implement AI. They have a budget for it, but it's not enough. They have no idea what it entails. The data isn’t ready and the foundations are not in place.
For example, in my most recent role with a start-up, the company gave me the impression that it had everything in place – the data scientists, the models and the servers. I questioned them pretty closely, but the company was hybrid and it was difficult to really gauge what was going on there. It was only when I accepted the job and went to an onsite meeting that I discovered I was the only data scientist in the building. They wanted me to build an entire LLM on a limited budget, without the servers, and with new models. It was impossible, which is why I no longer have that job.
I’m still looking for a new role. I’m sector agnostic – I’ve worked in finance, gaming, and start-ups, although I’m wary of big banks because of the long hours and the need to be in the office. It seems, though, that every job I apply for is deluged with applicants. There’s a continuous flow of new graduates, all of whom claim to be in AI or data science and who are competing for roles. Both data science and AI are meaningless PR terms: I’ve worked with so-called data scientists who don’t know how to code or how to use SQL.
The mass of candidates means that even finding a job is a problem. Most companies now use coding tests, which can be extremely onerous. They also take a long time to make hiring decisions: I’ve been in processes that have lasted months, and that’s a problem when you’re unemployed, as is unfortunately the case for me right now.
Have a confidential story, tip, or comment you’d like to share? Contact: email@example.com in the first instance. Whatsapp/Signal/Telegram also available (Telegram: @SarahButcher)
Bear with us if you leave a comment at the bottom of this article: all our comments are moderated by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. Eventually it will – unless it’s offensive or libelous (in which case it won’t.)