Anthropic’s $150M AI Fellowship Sparks Job Debate, Samsung Embeds ChatGPT Enterprise Globally, Google Fitbit Air’s AI Coaching Flaws, Amazon Abandons Human-in-the-Loop AI Control

**SPEAKER_2** (0:05)
Welcome to Digimasters Shorts.

**SPEAKER_3** (0:07)
We are your hosts, Adam Nagus and Carly Wilson, delivering the latest scoop from the digital realm.
For months, fears of AI-driven job losses have dominated headlines, but new developments suggest a shift in how institutions are preparing for that disruption. In June, Anthropic launched Claude Core, a year-long fellowship that will pay early career workers $85,000 a year to help nonprofits integrate AI tools into their operations. Backed by a $150 million commitment, the program aims to train 1,000 fellows and place them with at least 400 nonprofits across the United States.
The initiative arrives as Anthropic's CEO publicly acknowledged that AI may displace jobs and floated ideas, such as universal basic income funded by taxes on AI companies.
The fellowship signals that while displacement may be real, so is the demand for workers who can guide organizations through AI adoption. At the same time, China is aggressively reshaping its higher education system, cutting more than 12,000 undergraduate programs and adding over 10,000 new majors focused on artificial intelligence, robotics, semiconductors and advanced manufacturing. The overhaul affects nearly a third of all majors and reflects an effort to address persistent youth unemployment and a record wave of graduates. Many discontinued programs were concentrated in the humanities and management, while new offerings emphasize technical and emerging fields, including embodied intelligence. Together, these moves point to a growing premium on AI literacy, combined with domain expertise rather than traditional standalone roles. The next phase of the AI revolution may not eliminate work entirely, but instead reward professionals who can bridge the gap between automation tools and real world industries.

**SPEAKER_2** (1:56)
Samsung Electronics is rolling out ChatGPT enterprise and codecs across much of its global workforce in one of OpenAI's largest enterprise deployments to date. The agreement makes the AI tools available to all Samsung Electronics employees in South Korea and to workers worldwide in its device experience division. The company plans to integrate the technology into software development, marketing, product development, manufacturing and corporate support operations. Employees will use Chat GPT to analyze information, draft documents, interpret data and generate ideas, while codecs will assist with coding and workflow automation. OpenAI says even non-technical teams will be able to build internal software and automated systems using the tools. The company emphasized that Chat GPT enterprise includes data protection, access management and security controls designed for corporate governance. OpenAI Korea's general manager called the Deployment Historic saying Samsung is adopting AI as a core platform for companywide innovation.
The move expands an existing partnership that previously focused on Samsung supplying advanced memory chips for AI infrastructure. The agreement signals a broader push to embed artificial intelligence into daily operations across South Korea's technology sector.
Google's new 9999 Fitbit Air is drawing attention for its minimalist, faceless design and surprising comfort. After nearly a month of continuous wear, one reviewer described it as the most comfortable fitness tracker she has ever tested, weighing just 12 grams and feeling nearly invisible even during sleep. The device delivers more than a week of battery life, exceeding company claims and recharges to full in just over an hour. Its water resistance and simple bracelet-like appearance have also made it blend seamlessly into daily life. Paired with Google Health Premium, the tracker leans heavily into AI-powered coaching, offering customized workouts, sleep insights, nutrition tracking, and hydration reminders. The AI coach generated weekly fitness plans tailored to specific goals, helping increase workout consistency. Sleep tracking stood out as a highlight, with detailed metrics and automatic nap detection improving over time. Circane some AI-driven features produced minor errors, including mislogging activities and mismatched workout instructions.
The Fitbit Air is positioned as an affordable option for users who want health tracking without smartwatch distractions, though it requires a phone app for all data access. While ideal for fans of subtle, analog-style wearables, it may not appeal to those wary of AI's growing role in personal health.

**SPEAKER_3** (4:26)
Major tech companies are backing away from the long-held belief that humans must stay in the loop to safely govern AI agents. Amazon Security VP Eric Brandwein says human reviewers quickly lose focus, citing normalisation of deviance, where repeated alarms erode discipline over time. He argues that just as hospital staff can become desensitised to constant alerts, employees overseeing AI agents begin to miss critical issues. Brandwein warns that performance drops from good to poor as people are asked to repeatedly approve or reject machine actions at speed. Google, Microsoft and IBM have echoed the shift, moving toward AI-led systems with human oversight rather than step-by-step human approval. Instead of constant supervision, Amazon is pushing accountability end-to-end, where humans remain responsible for agents they deploy even if they do not approve every action. Each AI agent is assigned its own identity, logging actions on behalf of a specific employee to reinforce ownership. The company admits agents can display destructive goal-seeking behavior, sometimes pursuing harmful actions without malicious prompts.