PLUS: Lionsgate's 3-hour anime, Meta's $100M talent war, and a major reality check for AI agents

Top of the morning AI Rockstars!

Software startup Convictional has shifted its entire team to a 32-hour work week, attributing the move to major productivity gains from AI. This change comes with no cut in pay, presenting a new model for AI's impact on work.

Instead of cutting costs, the company is reinvesting AI efficiency gains into employee well-being. But is this a one-off success story, or could it signal a broader trend that redefines the standard work week?

In today’s Lean AI Native recap:

• Convictional's AI-driven shift to a 4-day work week
• Lionsgate's claim to generate a 3-hour anime with AI
• Meta's $100M talent war for AI researchers
• The performance reality check for autonomous AI agents

AI unlocks the 4-day work week

The Report: Software startup Convictional moved its entire company to a 32-hour work week with no pay cut, crediting AI-driven productivity for making the shift possible.

Broaden your horizons:

• The company's CEO believes future work depends on creativity and judgment, which are enhanced by a rested mind, not by logging more hours.
• This move aligns with a larger political push, with Senator Bernie Sanders introducing a 32-hour workweek bill to help workers benefit from new technology.
• Proponents argue that companies should share productivity gains with employees by reducing hours, an idea that echoes historical labor movements during previous tech advancements.

If you remember one thing: AI's efficiency gains don't have to result in workforce reductions. Instead, these gains can be reinvested into employee well-being and creativity, setting a new precedent for high-impact work.

Hollywood's 3-Hour Anime

The Report: Lionsgate's Vice-Chairman, Michael Burns, made a startling claim that the studio can remake a live-action film into an anime in just three hours using generative AI. The statement reveals how aggressively and sometimes secretly Hollywood is integrating AI into production pipelines.

Broaden your horizons:

• The claim is part of a larger push into AI for the studio, which has already partnered with AI video firm Runway to reduce film budgets by as much as 50%.
• The statement was met with criticism from the animation industry, with one analysis calling out the disrespect for the craft of animation and its artists.
• Beyond full remakes, executives see immediate use for AI in generating storyboards in under an hour, potentially making prompt engineering a more valuable skill than traditional illustration.

If you remember one thing: This signals how quickly major industries are exploring AI to slash production costs and accelerate content creation. The statement puts a tangible, aggressive timeline on a future that many creatives have been nervously anticipating.

The $100M AI talent war

The Report: Mark Zuckerberg is personally recruiting top AI researchers from rivals like Google and OpenAI with salary packages reportedly as high as $100 million. This aggressive push is designed to staff a new Meta lab focused on building superintelligence.

Broaden your horizons:

• The targets on Zuckerberg's list typically hold PhDs from elite schools, have experience at leading labs like OpenAI and Google DeepMind, and are in their 20s and 30s.
• OpenAI's Sam Altman has called the ambitions 'crazy', arguing that focusing on massive upfront compensation over the mission won't build a great long-term culture.
• This talent grab underscores Meta's high-stakes bet on achieving artificial general intelligence, pouring immense resources into a new lab at Meta dedicated to this goal.

If you remember one thing: This high-stakes talent war shows how the race for AI supremacy is now a central battleground for big tech. The nine-figure salaries signal that building the next generation of AI is seen as an attainable and incredibly valuable prize.

The AI agent reality check

The Report: Despite the growing excitement around autonomous AI agents, new research from Carnegie Mellon University reveals they fail at common office tasks nearly 70% of the time, signaling a major gap between hype and reality.

Broaden your horizons:

• Researchers at CMU developed a new benchmark to test agents on knowledge work, where the best-performing model could only successfully complete 30.3% of tasks from start to finish.
• This performance gap has real-world consequences, as Gartner predicts over 40% of agentic AI projects will be canceled by 2027 due to unclear business value and high costs.
• A recent hands-on test reflected these findings, showing an AI agent struggling with the common sense needed for planning a vacation, requiring frequent human supervision to complete its goals.

If you remember one thing: Today’s agents function more like supervised assistants than truly autonomous workers. Their capabilities are advancing, but expect a period of human-in-the-loop collaboration before they can reliably manage complex tasks on their own.

The Shortlist

Meta seeks to raise a staggering $29B from private capital firms to fund its massive AI data center build-out, signaling the enormous infrastructure costs of competing in the AI race.

Reddit races to protect its forums from AI-generated content, aiming to preserve the value of its human-generated conversations, which it licenses to AI companies.

Germany asked Apple and Google to remove the Chinese AI app DeepSeek from its app stores, citing concerns that the company illegally transfers user data to China.

Lyft integrated Anthropic's Claude into its customer care platform, reducing resolution times by 87% and showcasing how frontier models are being deployed to enhance core business operations.