AI Daily Briefing — 2026-05-27

At Anthropic’s recent Code with Claude event in London, developers were asked if they had shipped code written entirely by the AI. Nearly half raised their hands, with many admitting they hadn’t even reviewed the code before pushing it live. As tools like Claude Code grow more capable, a growing number of programmers are comfortable handing over their work to artificial intelligence. Anthropic says it aims to push automation as far as possible, but not everyone agrees this is a wise direction for the industry.

Meanwhile, the inaugural Enhanced Games kicks off this Sunday in Las Vegas, featuring 42 athletes competing while using performance-enhancing drugs. The event is designed to “push the boundaries of human performance” and reflects a broader cultural obsession with enhancement, from longevity optimization to extreme body modification. In 2026, the message seems clear: if you’re not enhancing, you’re falling behind.

At Google I/O, DeepMind CEO Demis Hassabis declared that we are “standing in the foothills of the singularity.” The company’s new Gemini for Science system leans into agentic, LLM-based AI that could eventually conduct cutting-edge research without human oversight. While specialized systems like WeatherNext still exist, Google appears to be shifting toward more autonomous AI. The move signals a major change in how AI will drive scientific discovery.

At Google I/O, DeepMind CEO Demis Hassabis declared humanity is at the “foothills of the singularity,” but the real story was the tension between grand visions and practical achievements. The keynote featured WeatherNext, an AI tool that provided early warnings for Hurricane Melissa in Jamaica, potentially saving lives. Yet this concrete success contrasted sharply with Hassabis’s talk of AI surpassing human intelligence, highlighting a growing divide in how AI is applied to science.

One approach focuses on specialized tools like WeatherNext or AlphaFold, designed to solve specific problems. The other envisions agentic, LLM-based systems that could conduct research autonomously. Google Cloud’s chief scientist recently argued that AI is moving beyond facilitating science to actually doing it. This shift raises questions about investing in narrow tools when broader, self-improving AI systems may soon take over.

Despite this, Google hasn’t abandoned specialized tools, releasing AlphaGenome and AlphaEarth last summer. AlphaFold remains widely used by over three million researchers, and Isomorphic Labs raised $2 billion for drug development. However, signs of realignment are emerging: Nobel winner John Jumper now works on AI coding, not science tools, as Google prioritizes agentic systems over specialized models.

Across the industry, autonomous research systems are gaining traction. While specialized tools still deliver real-world impact, the focus is shifting toward AI that can think and experiment independently. This could lead to a future where humans and AI collaborate as peers, or where AI drives scientific discovery on its own.

ClickUp CEO Zeb Evans has framed the company's recent 22% workforce reduction as a strategic pivot toward artificial intelligence, not a cost-cutting move. In a post on X, Evans announced that the collaboration software startup—valued at $4 billion in 2021—would redirect savings from the layoffs into million-dollar salary bands for remaining employees who deliver outsized impact using AI. The company has deployed roughly 3,000 internal AI agents to handle complex tasks, shifting staff roles from direct work to oversight and quality control.

ClickUp is not alone in betting on AI agents for productivity gains. A recent Gartner survey found that about 80% of companies using autonomous technology have cut jobs, though many have not seen corresponding financial returns. Critics warn that some firms use unproven AI as a pretext for downsizing, but Evans insists ClickUp is different, citing measurable efficiency improvements that the company plans to integrate into its customer-facing products.

Evans emphasized a shift from tracking AI token usage—a practice known as "tokenmaxxing"—to measuring value created and time saved. He argued that employees who automate their jobs with AI will remain indispensable, even as the company eventually needs fewer people. This vision mirrors a broader trend exemplified by Polsia, a one-person startup that recently raised $30 million at a $250 million valuation by handling all software operations through AI automation.

Despite widespread panic about artificial intelligence eliminating white-collar jobs, recent U.S. labor data suggests the technology has not yet caused mass unemployment. In fact, unemployment rates in occupations most exposed to AI are actually lower than in less-exposed fields. There is also no evidence of a major shift of workers from AI-threatened roles into manual-labor jobs, according to an analysis by MIT Technology Review AI. While the job market is undeniably struggling, the data indicates that AI is not the primary culprit behind current employment woes.

However, a more subtle disruption may be underway. A Stanford study found that young workers in AI-exposed occupations experienced a sharp decline in employment following the rise of generative AI. This trend was absent in low-exposure jobs, suggesting that AI is quietly replacing the junior-level tasks that traditionally serve as the first rung on the career ladder. This shift threatens to weaken entry-level opportunities for new graduates and early-career professionals.

The implications are significant. As AI takes over routine tasks once assigned to junior employees, businesses and educators must rethink how they train and support young people entering the workforce. Without intervention, the erosion of entry-level positions could create a lasting crisis for the next generation of workers, even as overall employment figures remain stable.

Despite widespread panic that artificial intelligence is wiping out white-collar jobs, economic data suggests the impact remains minimal. Analysis from the U.S. Bureau of Labor Statistics shows that unemployment rates for occupations most exposed to AI are actually lower than for less exposed roles. There is no evidence of a mass exodus from knowledge work into manual labor, challenging predictions of an imminent jobs apocalypse. While layoffs at tech giants like Coinbase and Meta fuel anxiety, the broader labor market remains stable, with AI disruptions still largely speculative.

Erika McEntarfer, a former BLS head now at Stanford, notes that only one in five U.S. companies use AI in any business function. "The data are a great reality check on the fear that AI will be enormously disruptive," she says. History shows innovations take time to transform industries and occupations. AI is unlikely to reshape labor markets until it first transforms businesses, which has not yet happened at scale.

The job market does remain tough for recent college graduates, with unemployment around 5.6%—a level not seen since the pandemic. Hiring rates have been dismal, especially for young people seeking tech roles. However, economists caution against blaming AI entirely. The current pain may stem from broader economic factors rather than automation, and the data suggests there is still time to plan for future disruptions before they arrive.