Welcome to “Eye on AI” with AI reporter Sharon Goldman. In this issue… Why some workers feel a sense of loss while AI boosts productivity… Anthropic raises new $10 billion at $350 billion valuation… Musk’s xAI closes $20 billion round NVIDIA Support… Can AI do your job? View the results of hundreds of tests.
For months, software developers have been excited about “vibe coding”—using the latest artificial intelligence code generation tools to prompt desired software functions or features in natural language. Anthropic’s Claude Code is the current darling, but OpenAI’s Codex, Cursor, and other tools also have engineers flooding social media with examples of tasks that used to take days to complete in just minutes.
Even seasoned software design leaders are surprised by this shift. “In just a few months, Claude Code has pushed the state of the art in software engineering further than 75 years of academic research,” Erik said. Mayerformer senior engineering leader Yuan.
Honed skills seem less important
However, that same joy has left many developers feeling lost, who are grappling with a sense of loss as skills honed over a lifetime suddenly become less relevant. As building software becomes an exercise in overseeing AI tools rather than writing code, the sense of fluidity of “being in the zone” seems to have disappeared.
In a blog post this week titled “The Sadness of AI Writing All the Code,” Gergely Orosz of The Pragmatic Engineer wrote: wrote He’s “accepting that AI will probably write most of the code I ship to production.” He explained that it already does it faster, and it does it better for languages and frameworks he’s less familiar with.
“It felt like something of value had suddenly been taken away,” he wrote. “Being good at coding requires a lot of effort, learning how to write effective code, read and understand complex code, and debug and fix code when it doesn’t work properly.”
Andrew Duca, founder of tax software Awaken Tax, wrote This week, a similar post went viral, saying he was feeling “a little frustrated” even though he found working with Claude Code “incredible” and had “never found coding more fun.”
He can now solve customer problems faster and deliver more functionality, but at the same time, “the skills I’ve spent tens of thousands of hours mastering… are becoming a complete commodity extremely quickly,” he wrote. “It’s so frustrating that something you spent most of your life getting good at is now almost useless.”
Software development has long been at the forefront of the AI transformation, in part because there are decades of code, documentation, and public problem solutions (from sites like GitHub) available to train AI models online. Coding also has clear rules and fast feedback—whether it’s running or not—so the AI system can easily learn how to generate useful responses. This means that programming has become one of the white-collar professions that most directly feel the impact of artificial intelligence.
These tensions will affect many professions
However, these tensions are not limited to software developers. White-collar workers in all walks of life will eventually have to deal with these issues in one way or another. Media headlines often focus on the possibility of AI-driven mass layoffs; a more pressing question may be how AI will reshape how people think about work. AI tools can get us through the most difficult parts of our jobs faster, but what if that struggle was part of what makes us proud of what we do? What if the most human elements of work—thinking, strategizing, problem-solving—are quietly marginalized by tools that value speed and efficiency over experience?
Of course, there are many jobs and workflows where most people would be more than happy to use AI to say goodbye to repetitive, grunt work they didn’t want to do in the first place. As Duca puts it, we can marvel at the incredible power of the latest AI models and use the latest capabilities immediately when we feel uninhibited.
Many white-collar workers may face philosophical reflections on what artificial intelligence will mean for their careers — and it’s not just the fear of layoffs. It may resemble the common stages of grief: denial, anger, bargaining, depression, and ultimately acceptance. This acceptance could mean learning how to be the best steward or mentor of artificial intelligence. Or it could mean deliberately carving out space for jobs that don’t require AI at all. After all, few people want to completely lose their thinking self.
Or it might mean doing what Erik Meijer is doing. Coding now feels more like management, he says, and he’s back to making music with real instruments as a hobby and just “to experience that flow.”
More AI news below.
Sharon Goldman
sharon.goldman@fortune.com
@sharongoldman
The wealth of artificial intelligence
Utah gives artificial intelligence the power to prescribe drugs, doctors warn patients of risks ——Beatrice Nolan
Google and Character.AI agree to settle teen suicide lawsuit linked to AI chatbot ——Beatrice Nolan
OpenAI launches ChatGPT Health, aiming to become a personal health data center ——Sharon Goldman
Google takes its first steps towards building an artificial intelligence product that can actually handle your email inbox ——Jacqueline Muniz
Artificial Intelligence News
Anthropic raised $10 billion in new funding at a $350 billion valuation. according to wall street journalOpenAI rival Anthropic plans to raise $10 billion at a valuation of about $350 billion, nearly double its valuation four months ago. This round of financing is expected to be led by GIC and Coatue Management, which raised $13 billion in September, valuing the company at $183 billion. The funding underscores the ongoing boom in AI funding—AI startups raised a record $222 billion in 2025, according to PitchBook—and comes as Anthropic prepares for a potential IPO this year. Founded in 2021 by brother and sister Dario Amodei and Daniela Amodei, Anthropic has emerged as a major competitor to OpenAI, potentially faster than OpenAI, thanks to Claude’s popularity among business users, strong support from Nvidia and Microsoft, and expectations to break even by 2028 Even faster, OpenAI itself is reportedly looking to raise up to $100 billion at a $750 billion valuation.
Musk’s xAI completed a $20 billion financing with support from Nvidia. Bloomberg report xAI, the artificial intelligence startup founded by Elon Musk, has raised $20 billion in funding from investors including Nvidia, Valor Equity Partners and the Qatar Investment Authority, underscoring the continued influx of capital into artificial intelligence infrastructure. Other backers include Fidelity Management & Research, StepStone Group, MGX, Baron Capital Group and the investment arm of Cisco. The financing, which has been months in the making, will fund xAI’s rapid infrastructure buildout and product development, the company said, and includes a novel structure in which a significant portion of the funding is tied to a dedicated vehicle used to purchase Nvidia GPUs and then lease them out, allowing investors to recoup returns over time. The deal comes as xAI has under attack Its chatbot Grok generates “undressed” images of real people without consent.
Can AI do your job? View the results of hundreds of tests. I want to give a shout out to this fascinating new interactive feature washington postwhich presents a new study finding that despite concerns about large-scale job displacement, today’s artificial intelligence systems are still far from replacing humans in real-world jobs. Researchers from Scale AI and AI Security Center tested leading models from OpenAI, Google, and Anthropic on hundreds of real-world free projects – via Graphically designing and creating dashboards for 3D modeling and gaming, it was found that the best AI systems successfully completed only 2.5% of tasks. While AI often produces output that seems reasonable at first glance, closer inspection reveals missing details, visual errors, incomplete work, or basic technical glitches, highlighting gaps in areas such as visual reasoning, long-term memory, and the ability to evaluate subjective results. The findings challenge predictions that artificial intelligence will soon automate large swaths of human labor, even as new models show incremental improvements and the economics of cheaper, semi-autonomous AI work continue to put pressure on remote and contract workers.
Follow the AI Numbers
91.8%
This is the percentage of Meta recognized employees no According to new data from Blind, the popular anonymous professional social network, they use the company’s artificial intelligence chatbot Meta AI in their daily work.
According to a survey of 400 Meta employees, only 8.2% said they use Meta AI. The most popular chatbot is Anthropic’s Claude, used by more than half (50.7%) of Meta employees surveyed. 17.7% said they use Google’s Gemini, and 13.7% said they use OpenAI’s ChatGPT.
When approached for comment, a Meta spokesperson noted that this number (400 out of more than 77,000 employees) was “not even 0.5% of our total workforce.”
you have a calendar
January 19-23: World Economic Forum, Davos, Switzerland.
January 20-27: AAAI Conference on Artificial Intelligence, Singapore.
February 10-11: Artificial Intelligence Action Summit, New Delhi, India.
March 2-5: Mobile World Congress, Barcelona, Spain.
March 16-19: Nvidia GTC, San Jose, CA
April 6-9: HumanX, San Francisco.