All news
Politics
Business
Technology
Incidents
Sports
Weather
Science
Entertainment
World
Real Estate
AI News
A Single Video Is Enough to Teach a Robot a Task It Has Never SeenPhoto by OsloMetX on Pexels
AI News

A Single Video Is Enough to Teach a Robot a Task It Has Never Seen

SAPO Tek12 September 2026 at 16:39

The company Skild AI has developed an innovative system that allows robots to learn new tasks through a simple video demonstration. This advancement represents a significant shift in how robots are programmed, eliminating the need to write specific code for each job.

With this system, a robot can execute tasks it has never performed before, learning solely from observing a demonstration video. The technology allows the machine to understand and replicate actions without traditional manual programming.

This innovation may render obsolete the current practice of programming robots individually for each new task. The traditional process of specific coding for each job could be replaced by a simpler and more accessible learning-by-demonstration method.

Skild AI's development comes at a time when the industry is seeking more efficient automation solutions. The ability to teach robots through videos can democratize access to robotics and accelerate the adoption of these technologies across various sectors.

Why this matters

This innovation can democratize the use of robots, allowing people without programming knowledge to train them for specific tasks. For readers in the region, this suggests that in the near future it will be easier to automate processes at home or at work, as long as one can record a video of the desired task.

About this summary

This is our short summary of a report published by SAPO Tek on 12 September 2026 at 16:39; the full text stays with the publisher. In our feed it sits under AI News. We currently carry 629 items in that section.

All items in this section →
Source

Related articles

Setor da IA pede travões antes de chegar a ponto de não retorno
AI News

AI sector calls for brakes before reaching point of no return

Influential artificial intelligence sectors are calling for a moratorium on technology development, warning that no laboratory sufficiently masters the technology to continue advancing at full speed. OpenAI's chief scientist wrote that the sector is not prepared for the risks it faces, while a company researcher publicly criticized the sector for recklessness. Sam Altman admitted being willing to slow development — but only if competitors do the same, underscoring the competitive nature of the problem.

Jornal de Negócios12/09/26, 17:00
AI News

"We need to slow down". Anthropic CEO urges AI companies to moderate the pace of model development

Dario Amodei, CEO of Anthropic, made a public appeal for AI companies to moderate the pace of development of their models, just days after an employee of the company resigned, considering that Anthropic was downplaying the risks that AI poses to humanity.

RTP Notícias12/09/26, 16:50
AI News

Why are the people building the most powerful AI so worried about what it could do?

The viral dismissal of an AI researcher from Anthropic has reignited debates about the risks of artificial intelligence, raising questions about the technological race in the industry. The article explores the concerns of those building the most powerful AI systems and questions whether the industry is advancing too fast without adequately considering the potential dangers for humanity.

Tuugo12/09/26, 16:19
25 dos maiores matemáticos do mundo alertam para a IA
AI News

25 of the world's greatest mathematicians warn about AI

Twenty-five of the world's greatest mathematicians signed an open letter warning about the risks of artificial intelligence in mathematics, considering that AI is no longer limited to answering questions or generating code, but is being used to solve complex mathematical problems. The letter highlights ethical and methodological concerns about the growing role of AI in mathematical research.

SAPO Notícias12/09/26, 15:00