AI Under Scrutiny: Unveiling Risks, Resistance, and Real-World Impact

Artificial intelligence is reshaping industries, sparking heated debates, and raising new ethical, environmental, and practical questions. From business leaders resisting the AI wave to the growing risks of deepfakes in politics, this press review explores the latest controversies, breakthroughs, and dilemmas at the intersection of technology, human judgment, and societal trust. Dive in to discover how AI is both transforming and challenging the worlds of work, communication, and decision-making.

Resistance to Artificial Intelligence: Voices Against the AI Wave

Sabine Zetteler, who runs a London-based communications agency, remains unconvinced about the value of using AI. She questions the point of consuming content generated by AI, stating, "Why would I bother to read something someone couldn't be bothered to write?" Zetteler emphasizes the lack of joy and meaning in replacing human effort with AI, especially when it comes to tasks like writing or even making business decisions that affect employees' livelihoods. She also raises concerns about the environmental impact of AI, referencing a Goldman Sachs report that estimates a ChatGPT query uses nearly 10 times as much electricity as a Google search query.

Florence Achery, owner of Yoga Retreats & More, shares similar reservations, particularly about the environmental cost and the contradiction between AI and her business's focus on human connection. Sierra Hansen from Seattle worries that reliance on AI could erode critical thinking skills, arguing that humans should organize their own lives rather than delegate such tasks to AI tools. Meanwhile, Jackie Adams, a digital marketing professional, initially resisted AI but eventually adopted it due to workplace pressures and the need to stay competitive. She now uses AI for copywriting and photo editing but remains uneasy about its growing influence.

"If you want to know why a decision is made, we will need humans. If we don't care about that, then we will probably use AI," says James Brusseau, a philosophy professor specializing in AI ethics at Pace University in New York.

AI's environmental impact is a significant concern for some business owners.
There is a perceived loss of human connection and critical thinking with increased AI adoption.
Workplace pressures are pushing reluctant professionals to integrate AI into their routines.

Source: BBC, "The people refusing to use AI"

Key Concerns	Details
Environmental Impact	ChatGPT query uses nearly 10x electricity of a Google search (Goldman Sachs)
Human Connection	AI seen as "soulless" by some business owners
Critical Thinking	Fears of diminished problem-solving skills

Summary: Resistance to AI is driven by environmental, ethical, and personal concerns, but workplace realities are making it increasingly difficult to opt out.

AI Hallucinations: Accuracy Issues Worsen with More Powerful Systems

According to The New York Times, the latest generation of AI systems, including those from OpenAI, Google, and DeepSeek, are producing more errors even as their mathematical and reasoning abilities improve. A notable incident involved an AI support bot for Cursor, a programming tool, which incorrectly informed customers of a non-existent policy change, leading to confusion and cancellations. Michael Truell, Cursor's CEO, clarified that the AI's response was entirely incorrect and not based on company policy.

Despite widespread adoption of AI bots for various tasks, there is still no reliable way to ensure these systems always provide accurate information. The phenomenon of AI "hallucinations," where systems generate false or made-up content, remains a significant challenge. In one test, newer AI systems exhibited hallucination rates as high as 79 percent. The underlying cause of these inaccuracies is not fully understood, as AI models are based on complex mathematical systems that cannot inherently distinguish between true and false information.

AI hallucinations can lead to real-world consequences, such as customer confusion and loss of trust.
Even the most advanced AI systems are not immune to generating false information.
There is currently no definitive solution to prevent AI hallucinations.

Source: The New York Times, "A.I. Hallucinations Are Getting Worse, Even as New Systems Become More Powerful"

Tested Hallucination Rate	Up to 79% in newer AI systems

Summary: As AI systems become more powerful, their tendency to generate inaccurate or fabricated information is increasing, posing challenges for both users and developers.

Roborock Saros Z70: Multimodal AI Revolutionizes Home Cleaning

Roborock is set to launch the Saros Z70, the world's first mass-produced robot vacuum cleaner equipped with an OmniGrip intelligent, foldable robotic arm. This device leverages multimodal AI, which integrates data from various sensors, including 3D point cloud data, RGB and infrared cameras, and Time-of-Flight (ToF) sensors, to navigate and adapt to changing home environments. The Saros Z70 can pick up objects as small as two square centimeters and is capable of distinguishing between different types of items, such as socks and crumpled paper.

The robot's AI-driven system continuously updates its map of the home in real time, allowing it to adapt to new furniture layouts and obstacles. Safety features include weight sensors that prevent the arm from lifting objects heavier than 300 grams, anti-pinch and pressure sensors, and spatial sensors to avoid collisions with children, pets, or furniture. The device is only 7.98 cm (3.14 inches) tall, enabling it to clean under sofas and other low furniture. It can also be voice-activated and used to monitor pets, sending pictures to users as needed.

"The Saros Z70 can help users grasp the future with intelligence," said Richard Chang, founder and CEO, Roborock. "It can be a user's home helper, companion, and assistant — whatever it needs to be."

Multimodal AI enables real-time adaptation to dynamic home environments.
Advanced sensors ensure both effective cleaning and user safety.
The device learns user habits and updates itself over-the-air for continuous improvement.

Source: Business Insider, "Multimodal AI makes this robot vacuum cleaner smart enough to pick up your socks"

Device Height	7.98 cm (3.14 inches)
Object Detection	As small as 2 sq cm
Weight Limit for Arm	300 grams

Summary: The Saros Z70 represents a significant leap in home robotics, combining advanced AI and sensor technology to deliver intelligent, adaptive, and safe cleaning solutions.

Marc Andreessen: AI Can Do Every Job—Except Venture Capital

Marc Andreessen, cofounder of Andreessen Horowitz and a prominent venture capitalist, asserts that while AI can potentially perform every job in the world, it cannot replace the unique, intuitive skills required in venture capital. On a recent a16z podcast, Andreessen argued that the art of picking winners among entrepreneurs is an intrinsically human skill, involving taste, intuition, and the ability to work closely with founders. He points out that even the best VCs only succeed with about two out of ten investments, suggesting that if the process were purely scientific, success rates would be much higher.

However, a recent SAP survey cited in the article found that 75% of C-level executives at billion-dollar companies believe AI already provides better business advice than their friends and colleagues, and 38% trust AI to make business decisions. The article also critiques the venture capital industry's track record, noting that VCs often fund familiar faces rather than truly disruptive innovators, a pattern that AI could potentially disrupt by identifying merit-based opportunities.

"You're not just funding them. You have to actually work with them to execute the entire project. That's art. That's not science. That's art. We would like it to be science, but, like, it's art." — Marc Andreessen

Andreessen believes venture capital requires irreplaceable human intuition and judgment.
Survey data suggests growing trust in AI for business decision-making among executives.
The article questions whether the VC industry's "art" is truly beyond AI's reach.

Source: Business Insider, "Marc Andreessen thinks AI can do every job in the world — except his"

VC Success Rate (per Andreessen)	2 out of 10
Executives Trusting AI for Business Decisions (SAP Survey)	38%
Executives Saying AI Gives Better Advice (SAP Survey)	75%

Summary: While Andreessen champions the irreplaceable human element in venture capital, industry trends and survey data indicate that AI is increasingly trusted for high-level business decisions.

UnitedHealth: 1,000 AI Use Cases Transforming Healthcare Operations

UnitedHealth Group has announced the deployment of 1,000 artificial intelligence applications across its insurance, health delivery, and pharmacy units. These AI tools are used for a variety of functions, including transcribing clinician visits, summarizing data, processing claims, and powering customer-facing chatbots. Additionally, 20,000 of the company's engineers utilize AI to write software, highlighting the technology's deep integration into the organization's operations.

The company has faced controversy over the use of AI in claims processing, but its chief digital and technology officer assures that AI alone will never be responsible for denying a claim. The broad application of AI at UnitedHealth demonstrates the technology's potential to streamline administrative tasks and improve efficiency in the healthcare sector.

AI is used for transcription, data summarization, claims processing, and customer service.
20,000 engineers at UnitedHealth use AI for software development.
AI will not independently deny insurance claims, according to company leadership.

Source: WSJ, "UnitedHealth Now Has 1,000 AI Use Cases, Including in Claims"

AI Applications in Production	1,000
Engineers Using AI	20,000

Summary: UnitedHealth's extensive use of AI is transforming healthcare administration, with a strong emphasis on maintaining human oversight in critical decisions.

AI-Generated Images: The Risks of Deepfakes in Political Communication

Former U.S. President Donald Trump recently shared an AI-generated image of himself as the Pope, describing it as a joke. However, communications experts warn that AI is a powerful tool that can easily blur the line between true and false images, making it ripe for abuse. The incident has sparked debate about the dangers of deepfakes and the potential for AI-generated content to mislead the public, especially in the context of political communication.

Experts emphasize that while AI-generated images can be used humorously, they also pose significant risks by enabling the creation of convincing but false representations. The blending of real and fabricated visuals can undermine trust and make it increasingly difficult for audiences to discern authenticity in media.

AI-generated images can be used for both humor and deception.
Experts caution that deepfakes threaten the integrity of political and public discourse.
The ability to blend true and false images is a growing concern in the digital age.

Source: Reuters, "Trump calls AI pope image a joke, but experts say it’s no laughing matter"

Summary: The use of AI to create deepfakes is raising alarms among experts, who warn of the technology's potential to erode trust and facilitate misinformation in politics and beyond.

Einschätzung der Redaktion

The persistent skepticism toward artificial intelligence, as expressed by business owners and professionals, highlights a crucial counterbalance to the rapid adoption of AI technologies. Concerns about environmental impact, loss of human connection, and the erosion of critical thinking are not only valid but essential for shaping responsible AI integration. The tension between workplace pressures and personal reservations underscores the need for transparent dialogue and ethical frameworks that prioritize both human values and sustainability. If these concerns are not addressed, there is a risk that AI adoption could outpace society’s ability to manage its broader consequences, potentially leading to diminished trust, increased environmental strain, and a weakening of essential human skills.

Wider societal debate and regulatory oversight are necessary to ensure AI serves collective interests.
Ethical and environmental considerations must be integrated into AI development and deployment.
Maintaining human agency and critical thinking should remain a priority in the age of automation.

Infobox: The resistance to AI adoption is a vital reminder that technological progress must be balanced with ethical, environmental, and human-centric considerations to ensure long-term societal benefit.

Sources: