You’ve probably heard CEOs talking about how AI is reshaping the workforce—sometimes in ways that sound straight out of a sci-fi script. Goldman Sachs’ CEO recently suggested that AI could take over writing IPO prospectuses. At Meta, executives have hinted that AI will reduce the need for entry-level coding jobs. And it’s not just talk—companies are already shifting hiring strategies based on what AI can automate.
This week, Anthropic (the company behind Claude) released its first Economic Impact Index, analyzing which professionals use its chatbot the most and for what tasks. The findings? AI isn’t outright replacing workers just yet, but it’s getting closer.
The Jobs Most at Risk, According to ChatGPT
Beyond Anthropic’s findings, OpenAI has also weighed in. Using its new research tool, Deep Research, AI researcher Min Choi analyzed the 20 jobs most likely to be automated. The post has already surpassed one million views. Among the most at-risk professions: tax experts, data entry clerks, telemarketers, accountants, and legal assistants.
Instead of speculating, recent data from Upwork provides a clearer picture of how AI is affecting the job market. Based on a study analysing freelancing jobs from November 2022 (one month before ChatGPT’s release) to February 2024, it’s clear to see which roles are declining the most:
- Writing jobs: Down 33%
- Translation jobs: Down 19%
- Customer service jobs: Down 16%
It’s no surprise that writing jobs took the biggest hit—AI-generated content is now widespread, and companies are replacing freelance writers with tools like ChatGPT. Similarly, translation jobs are declining as AI-powered language models become more advanced. Customer service is also suffering, as chatbots replace human agents for many inquiries.
What is more surprising is the decline in job postings for programmers: the number of coder jobs have dropped by 70% compared to their peak two years ago. There’s been ongoing discussion about the decline in demand for junior engineers. Skilled developers are now achieving much more with AI tools—work that once required entire teams just a few years ago.
Which Jobs Are Holding Strong (or Growing)?
The highest use of AI is in software engineering, accounting for 37.2% of discussions. Writing and editing follow at 10.3%, with scientific research in third place at 8.2%.
According to Anthropic, this is because AI is particularly suited for technical tasks, making programmers and content creators quicker to adopt it compared to other fields. Additionally, scientific and educational professions tend to use AI more than their relative share in the job market, as these fields require extensive data analysis, experimentation, and content creation.
Despite fears that AI would decimate creative and technical fields, some categories have actually seen an increase in job postings:
- Video editing/production: Up 39%
- Graphic design: Up 8%
- Web design: Up 10%
- Backend development: Up 6%
- Frontend development: Up 4%
This suggests that while AI is making content creation easier, it still requires human oversight to produce high-quality video, design, and development work. AI tools for these fields exist but aren’t polished enough to replace skilled professionals.
The #1 AI Job Right Now? Chatbots.
The biggest explosion in job postings? AI chatbot development—up 2000% since ChatGPT launched.
While many assumed AI would create more demand for machine learning engineers and data annotators, the reality is different. Companies aren’t racing to build their own LLMs—they’re integrating OpenAI APIs into their products to automate customer interactions. This suggests that in the near term, the most immediate job impact of AI will be in customer service automation rather than deep AI research.
Naval Ravikant, founder of AngelList has therefore given very clear advice for job seekers:
AI Agents are entering every role in the company
As AI Agents continue to improve in quality and popularity, I predict we will see more jobs affected by AI. A recent post by Sabrina Wu and Vivek Ramaswami of Madrona, shows what roles AI agents are already impacting in a company’s P&L:
- Taxes & Interest: Emerging AI solutions are optimizing tax management, sales tracking, and financial reporting.
- Cost of Goods Sold (COGS): AI-powered agents are enhancing margins by automating large-scale manual tasks. This includes site reliability co-pilots, cloud cost optimization tools, and AI-driven customer service solutions.
- Sales & Marketing (S&M): AI is revolutionizing sales with automated lead generation, intelligent CRMs, and AI-assisted deal-closing tools. In marketing, SEO is evolving to optimize for AI-driven search, while AI-powered branding adapts content dynamically.
- Research & Development (R&D): AI is accelerating product development, from AI-assisted coding and automated QA testing to dynamic UI/UX design and model training enhancements.
- General & Administrative (G&A): AI is transforming back-office functions by streamlining finance, HR, IT, and security—turning cost centers into strategic advantages.
These shifts align with broader labor market trends. The World Economic Forum’s Future of Jobs Report 2025 anticipates a net increase of 78 million jobs by 2030, despite automation displacing 92 million roles. Growth is expected in technology, green energy, and caregiving, reinforcing the need for reskilling, with 39% of core job skills requiring updates. As AI-driven transformation accelerates, businesses are prioritizing skills-based hiring and flexible work models to stay competitive.
Why This Matters
The data tells a clear story: AI is already reshaping the workforce, but it’s not a blanket replacement for all jobs. The roles disappearing first are those that AI can do quickly and cheaply (writing, translation, customer service). Meanwhile, jobs requiring creativity, technical expertise, or high-quality execution are still in demand—and in some cases, even increasing.
For professionals, the takeaway is clear: The best way to stay ahead isn’t to fight AI but to learn how to use it effectively. Whether it’s integrating AI into your workflow or specialising in areas where AI still falls short, the winners in this shift will be those who adapt.
All that being said, there’s no need to panic (at least not immediately). Anecdotally, a lot of these AI Agent tools suffer from same problems we have all encountered with LLMs such as hallucinations, variable quality and challenges adapting to case-specific context. Several of these companies have experienced high growth, but also high churn.
As the adage goes, your job will not be replaced by AI, but it might replaced by a human that knows how to use AI.
I'm a former general partner at google ventures, head of Google for Entrepreneurs in Europe and founding head of Campus London, Google's first physical hub for startups.
I'm also the founder of Techbikers, a non-profit bringing together the startup ecosystem on cycling challenges in support of Room to Read. Since inception in 2012 we've built 11 schools and 50 libraries in the developing world.
- Weekly Firgun Newsletter – March 20 2025 - March 21, 2025
- Don’t Sleep on Consumer Tech: AI is Changing Everything - March 16, 2025
- Weekly Firgun Newsletter – March 14 2025 - March 14, 2025
