fbpx
01 / 05
Anthropic Releases “Most Intelligent” AI Model in Rivalry with OpenAI

Bloomberg | Computing

Anthropic Releases “Most Intelligent” AI Model in Rivalry with OpenAI

“Anthropic introduced Claude 3.5 Sonnet, an updated version of its Claude 3 model that it just released in March. The company said Thursday that the new model — the technology that underpins its popular chatbot Claude — is twice as fast as its most powerful previous version. Anthropic said in its evaluations, the model outperforms leading competitors like OpenAI on several key intelligence capabilities, such as coding and text-based reasoning.”

From Bloomberg.

Wall Street Journal | Computing

New “Automated Reasoning” to Reduce AI’s Hallucinations

“Amazon is using math to help solve one of artificial intelligence’s most intractable problems: its tendency to make up answers, and to repeat them back to us with confidence.

The issue, known as hallucinations, have been a problem for users since AI chatbots hit the mainstream over two years ago. They’ve caused people and businesses to hesitate before trusting AI chatbots with important questions. And, they occur with any AI model—from those developed by OpenAI and Meta Platforms to those from the Chinese firm DeepSeek.

Now, Amazon.com’s cloud-computing unit is looking to ‘automated reasoning’ to provide hard, mathematical proof that AI models’ hallucinations can be stopped, at least in certain areas. By doing so, Amazon Web Services could unlock millions of dollars worth of AI deals with businesses, some analysts say.

Simply put, automated reasoning aims to use mathematical proof to assure that a system will or will not behave a certain way. It’s somewhat similar to the idea that AI models can ‘reason’ through problems, but in this case, it’s used to check that the models themselves are providing accurate answers.”

From Wall Street Journal.

CNN | Scientific Research

First Words Decoded in 2,000-Year-Old Charred Scroll

“Scholars are decoding an ancient scroll that was one of hundreds charred to a crisp during the eruption of Mount Vesuvius in AD 79.

The artifact, which is kept at the University of Oxford’s Bodleian Libraries in the United Kingdom, is the fifth intact Herculaneum scroll to be virtually unrolled as part of the Vesuvius Challenge, a competition designed to accelerate the deciphering of the scrolls that form an unprecedented cache of information about ancient Rome and Greece.

Using artificial intelligence and other computer-based techniques to piece together the scroll and enhance the ink, the Vesuvius Challenge team has successfully generated the first images of text inside the scroll, known as PHerc. 172. The library group said Wednesday that it has begun to interpret the columns of text. 

One of the first words to be translated was the ancient Greek διατροπή, meaning ‘disgust,’ which appears twice within a few columns of text, the Bodleian Libraries said.”

From CNN.

Live Science | Science & Technology

New AI-Designed Material Is as Light as Foam but Strong as Steel

“Scientists have used artificial intelligence (AI) to design never-before-seen nanomaterials with the strength of carbon steel and the lightness of styrofoam.

The new nanomaterials, made using machine learning and a 3D printer, more than doubled the strength of existing designs. The scientists behind the new study said they could be used in stronger, lighter and more fuel-efficient components for airplanes and cars. They published their findings Jan. 23 in the journal Advanced Materials.”

From Live Science.

World Bank | Adoption of Technology

Senegal Closing on Universal Electricity Access

“Khady’s story reflects Senegal’s broader journey toward universal electricity access. The country has made significant strides, with a national access rate of 84% according to government sources. This progress has transformed millions of lives, yet the journey is far from over. While urban areas enjoy near-universal access, over 30 % of rural communities remain disconnected from the grid.”

From World Bank.