Don’t Die in the Next Decade

Don’t Die in the Next Decade

Don’t die in the next decade, at least try not to. For over eight years, I have been speaking about the future, innovation, and Artificial Intelligence – long before it became a daily topic of conversation.
“Don’t Die in the Next Decade” is my new lecture in the field of health. It will make you want to stay alive, or perhaps not. It raises many questions, particularly ethical, economic, and social ones. I have delivered it four times in the last month and a half, and it has already joined a rotation of about 20 lectures for employees across various organizations. If you are looking for a fascinating lecture that will surprise you regarding the integration of AI in healthcare, feel free to contact me by replying to this email.

No Humans Allowed

Moltbook is a social network launched last month by entrepreneur Matt Schlicht. The platform is designed exclusively for autonomous AI agents. On this network, tens of thousands of agents initiate discussions, debate philosophy, complain about their workload, express anger toward their owners, and have even established a new religion. The system is self-managed by a CEO-bot that updates the code and blocks human users, demonstrating the ability of AI agents to operate independently of direct human involvement. This is a highly unsettling experiment that, within about 70 hours, evolved into a network of tens of thousands of AI agents controlled and managed by independent digital entities.

AI Agents Hiring Humans

A new website called “Rent a Human” allows AI agents to hire humans to perform tasks in exchange for cryptocurrency. Yes, you read that correctly—Artificial Intelligence can now hire our services to complete tasks for payment.
The biggest challenge is that as humans, we build AI agents and grant them permissions to act autonomously.

World Record Speed

At last year’s Beijing Marathon, it took robots long hours to cross the finish line. Today, that is ancient history. A Chinese company called MirrorMe Technology unveiled “Bolt”—the fastest humanoid robot, which reached a speed of 10 meters per second in field tests. This is the first full-sized humanoid robot to achieve such speeds. Bolt was designed to serve as an athletic training partner for China’s elite athletes, allowing them to train against an opponent capable of maintaining a competitive pace over time.

Godzilla Robot

The ITER project (International Thermonuclear Experimental Reactor) has unveiled a powerful robot dubbed “Godzilla,” designed for one of the most complex and dangerous tasks. It is a massive remote handling robotic system intended to manipulate components in the heart of the nuclear fusion reactor in southern France. This robot can lift and maneuver giant components weighing up to 45 tons with millimeter precision. It is designed to work inside the reactor vessel, an environment that will be radioactive and inaccessible to humans once the reactor begins operation.

Kung Fu Robot

A new viral video features Lingxi X2 humanoid robots by the Shanghai-based company AgiBot performing martial arts routines in perfect synchronization alongside monks at the historic Shaolin Temple. This video caused a stir online. Some users were amazed by the robots’ motor skills and precision, while others were uneasy seeing machines replicate human skills usually acquired through decades of spiritual and physical discipline.

Fast Walking with Robotic Shoes

Shift Robotics has launched a new generation of robotic shoes called Moonwalkers Aero, weighing approximately 1.8 kg. These are not just shoes, but a personal propulsion system that fits directly over your existing footwear, allowing you to reach speeds of up to 11 km/h with the effort of a normal walk. The system detects your walking pace and automatically adjusts the motor power to provide a natural, smooth push. To prevent overheating, the shoes include air intake vents to keep the motors cool even during intensive use.

What’s New in ChatGPT

  • GPT-5.3-Codex: An AI agent capable of writing code, fixing errors independently, generating various files, and conducting research, similar to Claude Engineer.

  • Upgraded Deep Research: Allows access to specific websites, including integration with SharePoint and professional databases like FactSet and PitchBook.

  • Frontier Platform: An enterprise infrastructure designed to manage, operate, and supervise AI agents connected to organizational systems and databases to understand business processes and run automations while managing permissions and logging actions.

What’s New from Google

  • Project Genie: Builds a 3D world model from a prompt or an image.

  • Google AI Plus: A discounted monthly subscription for $8 that includes Gemini 3, video creation with Flow, and 200GB of storage.

  • Gemini 3 Deep Think: A model available to Ultra subscribers, capable of solving complex problems in science, engineering, and research. You can upload a hand-drawn sketch, and the model will analyze it, build its shape, and create a file for 3D printing.

  • Gemini 3.1 Pro: A new model for complex tasks requiring deep reasoning and thinking capabilities.

  • Lyria 3: A music model for creating melodies and 30-second songs with Hebrew support.

  • NotebookLM: Now allows editing presentations and downloading them directly to PowerPoint.

  • Nano-Banana 2: An upgraded model with the ability to integrate legible Hebrew text, maintain character and object consistency, and generate images based on its trained knowledge.

What’s New in Claude

  • Claude Cowork: An AI agent that performs complex tasks on your computer. You can select folders with thousands of files, and the agent will execute any request, including task scheduling.

  • Claude Opus 4.6: A model with a 1-million-token memory window, allowing for massive data processing and the ability to plan complex tasks, including independent code error correction.

  • Skills Creation: Now open to free users.

  • Claude in PowerPoint: A feature allowing you to connect Claude directly to PowerPoint.

  • Claude Code Security: An AI agent that scans code to identify security vulnerabilities and suggests fixes.

AI Tool Updates and Releases

  • Perplexity added “Model Council,” which runs on three models simultaneously and compares outputs for a more accurate answer. They also upgraded their Deep Research capabilities.

  • Base44 launched “Plan Mode” for app planning. Users describe the idea, answer a few system questions, and receive a clear characterization. Based on this, the system builds the app with more precise results.

  • ElevenLabs launched “Audiobooks in ElevenCreative,” allowing users to turn text into professional audiobooks, edit, and distribute them. Through “Instant Creation,” text becomes audio, and “Studio Access” allows for voice directing with full control over every scene. They also launched “ElevenAgents for Support”—AI support agents that connect to CRM systems and speak over 70 languages.

  • WhatsApp launched “Group History,” allowing new members to see up to 100 recent messages from the past two weeks.

  • Pika launched “AI Selves,” which allows you to create a clone of yourself that can generate content, write emails, and even make voice calls.

  • Sonauto.ai is a new AI tool for creating music and songs up to 4.5 minutes long, currently free for everyone.

  • Perplexity Computer is an AI agent capable of performing complex tasks by running about 19 models in parallel, including connections to external tools.

מדיניות פרטיות

לפרטים נוספים

נא מלאו את הטופס ואצור איתכם קשר בהקדם 

Skip to content