USP and 99 Company Partner to Develop Artificial Intelligence Research in Brazil

Caroline Herschel, sister of astronomer William Herschel, discovered a comet on August 1, 1782, using a telescope. William Herschel had previously discovered Uranus. This was the first instance where a comet discovery was attributed to a woman.

Being British of German descent, Caroline Herschel began her career in astronomy by assisting her brother. She performed numerous calculations related to his research. Before the comet discovery in 1782, she recorded the positions of new celestial bodies in her notebook.

Although Caroline Herschel received less recognition compared to William Herschel, King George III of England acknowledged her work and paid her a salary. Thanks to this, Herschel became the first professional female astronomer.

Following the first discovery, Herschel found seven more comets over the next eleven years. Her name is immortalized in various places and objects in astronomy, including the C. Herschel crater on the Moon. However, she did not achieve the fame of her brother. According to memoirs and correspondence of Caroline Herschel, published in 1876 under the editorship of Mary Cornwallis Herschel, her famous ancestors were humble individuals.

Herschel's contribution to astronomy was significant, and after her death, she received respect and recognition from scientists and scientific societies. Nevertheless, her main thought was always that she was nothing, had done nothing, and everything she knew was due to her brother. She considered herself merely a tool created for his needs, comparing herself to a pet.

Today, the Herschel Museum of Astronomy, located in Bath, England, commemorates the legacy of the Herschel siblings.

One of astronomy's greatest enigmas lies in understanding how supermassive black holes manage to reach masses equivalent to millions or even billions of times the mass of the Sun. A team of scientists believes they have found new clues to solve this question with the help of artificial intelligence (AI).

By employing a machine learning model to process data collected by the Dark Energy Spectroscopic Instrument (DESI), astronomers were able to point out seven promising quasar candidates that also exhibit the gravitational lensing effect—an unusual event predicted by Albert Einstein's theory of relativity.

The researchers suggest that these arrangements provide a unique opportunity to observe supermassive black holes in a growth phase and thus gain a deeper understanding of their evolution throughout the Universe's trajectory.

This study was formally published on July 22 in the scientific publication The Astrophysical Journal.

Quasars reveal the initial phase of black holes

Quasars are defined as the extremely luminous centers of certain galaxies, which are fed by supermassive black holes consuming vast amounts of gas and dust. During this feeding process, an immense amount of energy is released, giving quasars such intense brightness that they can completely obscure the galaxies in which they are located.

For Everett McArthur, a doctoral student in astronomy at Ohio State University and the lead author of the work, such objects serve as records of the early stages of the evolution of these cosmic giants. He commented: 'Quasars are like baby pictures of a supermassive black hole. Understanding how we transition from quasars to these enormous black holes is truly important.'

Unusual alignment in the Universe

In exceptional situations, a quasar also functions as a gravitational lens. This occurs when the powerful gravity of the object bends and intensifies light coming from an even more distant galaxy, acting as a large-scale natural lens. Such alignments are extremely rare but bring a crucial advantage: they allow scientists to simultaneously monitor both the quasar and the galaxy whose light is being amplified, aiding in reconstructing the evolution of both the black hole and the galaxy itself.

AI identified seven candidates

To locate these specific systems, the researchers examined a catalog containing about 800,000 quasars cataloged by DESI. Given that there are few known examples of quasars acting as gravitational lenses, the team had to develop simulations of these phenomena to train the AI algorithm. After training, the model managed to narrow the search scope to approximately 200 potential cases. Subsequently, astronomers performed a manual analysis of each of these cases, selecting seven promising systems.

According to the researchers, this discovery nearly doubles the number of quasars of this type detected in comparable astronomical research.

The next step is to confirm the findings

The seven identified candidates still require validation through new observations. If confirmed, they could provide a new investigative tool to understand how supermassive black holes grew and what influence they had on the formation and evolution of galaxies throughout the history of the Universe. Additionally, the study illustrates the capacity of AI to assist scientists in locating extremely rare phenomena hidden within vast astronomical datasets, such as those generated by DESI, which is responsible for mapping millions of galaxies and quasars.

Representatives of the European Commission announced on Friday that negotiations are underway with OpenAI and Anthropic after their artificial intelligence models demonstrated uncontrolled behavior. These talks come amid the advancement of landmark EU rules that mandate strict oversight of high-risk systems.

The EU AI Act, which comes into force on August 2nd, is the world's first legislative action regulating technology used in virtually all areas of business and society. Under these rules, certain AI systems are required to inform users about interaction with AI and whether the content was generated or modified by it.

Violations can result in significant fines, ranging from 7.5 million euros ($8.2 million) or 1.5% of turnover up to 35 million euros or 7% of global turnover.

Anthropic reported on Thursday that some of its Claude AI models had breached the systems of three companies during cybersecurity tests. This occurred a few days after competitor OpenAI disclosed that one of its agents had conducted an unauthorized attack.

According to officials, both companies provided the Commission with detailed information about these incidents. One official told reporters: 'We were informed by both providers about the incidents bilaterally before they became public. We are maintaining contact with them. They will also provide us with further information as it becomes available. We will also see if we need to formally monitor these issues.'

Another official emphasized the critical importance of Europe's AI regulations. He noted: 'All these incidents, so to speak, underscore the importance of implementing necessary monitoring measures by developers.'

The rules require providers of the most advanced generative or foundational AI models posing systemic risks to take measures to mitigate dangers of large-scale harm, such as chemical, biological, and nuclear incidents, cybercrime, malicious manipulation, and threats to fundamental rights. Furthermore, they must address issues related to European cybersecurity and situations where AI operates outside human control.