In 2023, the tech industry saw waves of layoffs, which will likely continue into 2024.

Due to the rise of LLMs and the shift towards pre-trained models and prompt engineering, specialists in traditional NLP approaches are particularly at risk.

Neither teams working on proof-of-concept projects nor production ML systems are immune from job cuts.

Data scientists and NLP specialists can move towards analytical roles or into engineering to stay relevant. In any case, they should hone their essential communication, business, and technical skills.

If Oxford declared that the word of the year for 2023 was ‘layoff,’ it wouldn’t surprise tens of thousands of people across the globe. In a time where economic challenges force companies to streamline operations, machine-learning (ML) specialists and adjacent roles are not immune to the trend of mass layoffs.

The rapid advancements of Large Language Models (LLMs) are changing the day-to-day work of ML practitioners and how company leadership thinks about AI. Are LLMs entirely overtaking AI and natural language processing (NLP)? Could this paradigm shift lead to widespread job reductions? Who are the people most at risk of being laid off?

Piotr Niedźwiedź, Founder and CEO of, and I discussed this and more in our 2023 Year in Review episode of the ML Platform Podcast. Let’s recap some key points.

The rise of modern LLMs

In 2023, the dominance of modern LLMs became increasingly evident, challenging the incumbent classical NLP models. Even small and relatively weaker LLMs like DistilGPT2 and t5-small have surpassed classical NLP models in understanding context and generating coherent text. Anyone with a stable internet connection can feed a text to an LLM and get a comprehensive summary, extract answers from it, or have it rewritten.

As pre-trained models are prevalent and fine-tuning is increasingly replaced by prompting, machine-learning and even software engineers can now manage sophisticated NLP setups without the support of specialized data scientists.

This development leaves those data scientists in a tough spot: Will their NLP skills still be relevant to employers in a couple of years? Or should they start to look for new career opportunities?

The lifecycle of NLP projects: PoCs and production

As the tech industry faces waves of layoffs, it’s worth understanding the dynamics of the NLP project lifecycle to assess the risk of job cuts for those involved.

We believe it’s instructive to differentiate between NLP projects already in production and those in the proof-of-concept (PoC) stage.

PoC projects are trial runs, aiming to prove the worth of a new technology to a business. They often do not show tangible results right away, making the people working on them seem expendable. That’s particularly true in times when managers quickly cut projects without an immediate, clearly measurable impact on the bottom line. However, C-level executives might find it easier to justify spending on trendy GenAI solutions to their investors than obtaining buy-in for attempts to revive a struggling product line.

NLP projects in production face their own set of challenges. With the rise of LLMs, teams running applications on more traditional NLP approaches must decide whether to continue investing in their current stack or switch to LLMs. This decision impacts both jobs and project continuity. For specialists immersed in these projects, there is growing uncertainty about their projects’ direction.

As you can see, it’s unclear whether people working on PoC or production projects are at higher risk of layoffs. As Piotr warns, there’s a lot of gray area, and we agreed that it’s too soon to tell how large of an impact the rise of LLMs will have on global tech layoffs.

Evolve or sink

So, where do we go from here? There is no straightforward roadmap, but those at risk of layoffs should adjust to the situation instead of letting it dictate their course. Data scientists may need to transform their roles to flip the narrative. One possible transformation is shifting towards business intelligence (BI) or business analytics roles by embracing their analytical skills.

Another possibility is to move towards software engineering. We’re already witnessing a rise in engineers who don’t consider themselves machine learning engineers but work with ML technology daily.

Irrespective of the direction you want to take, honing some fundamental skills is always a good idea to shield yourself from layoffs as much as possible. These include:

  • 1
    Written and oral communication: Practice effectively communicating technical solutions and analytical insights to colleagues and non-technical stakeholders.
  • 2
    Business proficiency: Learn to understand and communicate your work’s impact on the business’s overall success. In economically challenging times, management values employees who know how to prioritize and identify cost-cutting opportunities.
  • 3
    Continuous learning and professional development: Stay updated with the latest advancements by attending conferences, participating in online courses, and actively engaging with the community through forums and meetups.

Predictions and considerations for the future

As the discussion drew to a close, both Piotr and I agreed on a few key takeaways about the current landscape of ML layoffs:

  • The rise of LLMs is undeniable, challenging the established roles of classical NLP models, but the full-scale replacement of traditional NLP models might take longer than some anticipate.
  • Global economic needs, efficiency requirements, and the distinction between value-proven production systems and experimental PoC projects will likely play significant roles in shaping the future trajectory of machine-learning careers.
  • Projects still in the PoC stage are at higher risk of being cut, while those already in use must decide whether to incorporate LLMs or further invest in their current tech stack.
  • Professionals from various fields are all waiting at the edge of their seats to see how the AI revolution pans out. In the meantime, they may need to diversify their skill sets to stay essential amidst job cuts. 

You can watch the full episode on YouTube:

You can also find the ML platform podcast on all your favorite podcast streaming platforms:

Was the article useful?

Thank you for your feedback!

Explore more content topics: