FX.co ★ Beyond singularity: tomorrow arrives today

Can AI recruiters be trusted?

Altman emphasized that boosted AI productivity will force people to work even more. Paradoxically, AI agents themselves now increasingly handle hiring. Neural networks read resumes and conduct interviews, analyzing online facial micro‑expressions, voice timbre, stress resilience, and the logic of a candidate’s answers. An AI recruiter can independently decide to hire, generate an offer, and sign a contract without human involvement. This speeds hiring by orders of magnitude but raises questions about the impartiality of algorithmic judgments.

Markdown — ads for bots

Everyday web search has also changed forever: advertisers began to shift from people to AI agents. Time and other media giants started converting web pages from traditional HTML into plain Markdown without design or images. The idea is that AI assistants can instantly parse content and embed brand advertising directly into their replies to users, bypassing classic browser banners.

Invisible tags from Anthropic

Those who hoped to ascend to the creative Olympus with AI will be disappointed. Anthropic released a tool that undermines ambitious plans to publish novels generated by neural networks. New Claude models automatically inject an invisible watermark into generated text that remains even after human editing. The innovation was implemented to satisfy stringent EU AI‑transparency requirements. Anthropic will provide third‑party developers with special software to detect these hidden marks.

Cybercab + Starlink — Tesla robotaxi

Changes are coming to the taxi market. Tesla unveiled a production version of its fully driverless robotaxi, Cybercab. Its main technological feature is a Starlink V5 satellite Internet terminal integrated into the roof at the factory. Elon Musk said high‑speed space Internet will become a baseline standard and will be integrated into all new cars from the brand. That will allow robotaxis to remain in constant contact with a unified computing center, processing massive road‑data streams in real time.

Enigma — robots study how we control them

Israeli startup Enigma raised $71 million for a large‑scale experiment. The company granted remote access to 100 physical robots to users around the world. Instead of training AI right away, developers first studied people to find how humans prefer to control machines—by voice, gestures, or video commands. The collected data will help create intuitive interfaces and lay the foundation for a new generation of robotic AI that understands human habits.

Claude thinks aloud

While Israeli robots learn, American AI agents expand both their own and human capacities. Anthropic advanced its voice mode for the Claude assistant. Voice interaction now works with flagship Opus and Sonnet models. Previously, the voice was limited to the lightweight Haiku model for simple conversations. Users can now hold deep, complex architectural and scientific discussions by live voice while the neural network verbalizes its reasoning in real time.

Lunavity —drone backpack

The rotor drone‑backpack Lunavity also extends human horizons. Developed at the University of Tokyo, it is designed to let a person jump higher. Powerful propellers generate vertical thrust, allowing the user to leap two to three times higher than normal and land smoothly. The system recognizes owner gestures and movements, enabling direction changes mid‑air. Lunavity is planned for entertainment and for the rehabilitation of people with musculoskeletal impairments.

3D‑printed drugs tailored to your DNA

The pharmaceutical sector is undergoing a revolution thanks to home personalized drug synthesizers. Companies are testing compact devices that, based on continuous data from smartwatches and a patient’s genetic passport, print the required pills right in the kitchen. The printer mixes exact dosages of active ingredients tailored to the current state of the organism—from blood sugar to stress‑hormone levels—eliminating medical dosing errors and drug counterfeiting.

Cost of high IQ and tall stature

Offers from US startups like Nucleus, Orchid, and Herasight are already going beyond medicine. Using polygenic tests (PGT‑P), prospective parents can screen embryos and predict traits such as intelligence, height, lifespan, and eye color. Scientists debate the accuracy (IQ gains may amount to only a few points), and society fears a return to eugenics, yet investments in fertility tech are rising. For large sums, the line between treating disease and creating “enhanced” humans is rapidly blurring.