This issue is somewhat themed, with three recent articles looking at the impact of AI on work and the job markets. Enjoy!

Remote Labor Index: Measuring AI Automation of Remote Work

REMOTELABOR.AI

The Remote Labor Index indicates that AI is a long way from replacing human workers - with 2.5% task completion rate.

Most benchmarks are based on small tasks or academic-style tests that are not a very good reflection of our day-to-day work. Recently we’ve seen researchers try to create more representative tests, including OpenAI’s GDPval, which used a group of experts to create representative tasks across a range of professions.

This paper introduces the Remote Labor Index, similar to SWE-Bench (which focusses specifically on software engineering), it is built from a collection of real-world tasks. In this case, 240 projects from upwork, which had a human completion time of ~13 hours. This are sizeable, highly involved tasks, with clear economic value.

The result? The leading AI models were able successfully complete just 2.5% of tasks to a sufficient level of quality.

While this might sound like a very poor result, personally I am deeply impressed that general purpose AI models are able to successfully complete highly-involved tasks, with genuine business value (worth 100s of dollars) at all. This was impossible just a year or two ago.

Furthermore, the AI models completed a far higher percentage of tasks, but the quality was deemed insufficient.

We are now at a point where general purpose AI models are able to complete a small percentage of randomly selected, economically valuable tasks, across a range of domains.

I’m impressed.

I analyzed 180M jobs to see what jobs AI is actually replacing today

BLOOMBERRY.COM

This blog post is based on analysis of global job openings, by job title, comparing 2025 data to 2024. The results don’t necessarily prove a connection between AI and shifts in job openings, but the trends are interesting nonetheless.

Within software engineering, the overall number of job openings haven’t changed much in the last year. However, we are seeing a significant shift in the types of role:

There is a modest drop in front-end and mobile roles, whereas data engineering and machine learning especially are showing significant increases. It is probably safe to say that this surge is as a direct result on the growing interest in AI.

AI Is Making It Harder for Junior Developers to Get Hired

FINALROUNDAI.COM

This blog post has a more nuanced take on AIs impact within the job market, with much of it based on the headline results of a recent Harvard study.

The researchers looked at resume and job postings on LinkedIn, and used a novel GenAI powered method to identify organisations they consider to be “GenAI adopters”. They conclude that:

“GenAI adoption coincides with a pronounced decline in junior employment”

In other words, organisations (across a range of sectors) that adopt AI are hiring fewer junior roles.

This blog post explores the more long-term impact of this change. Here is a standout quote from their commentary:

“Juniors have always been more than cheap labor. They were an investment. Companies hired them not because they were immediately productive, but because they grew fast and carried the culture forward. “