onesense onesense

Right now, in the tech industry paddock, there are 372 AI unicorns, according to PitchBook.

How did you react to that number? Did you double-take? Did you recoil slightly, finding it shockingly high? Far too low? Or did you shrug as if to say "that tracks"?

I don’t think there’s a right or wrong response here—my point, rather, is that it’s likely you’ll have some kind of reaction to that number. And that’s representative of the extent to which the AI boom has created so many undercurrents of optimism and skepticism. 

"The AI investing landscape is very competitive at the moment," said Kyle Stanford, PitchBook director of research for U.S. venture, via email. "We are seeing around 1/3rd of completed deals go to companies in the AI space across all industries, so that encompasses much more of the market than just the handful of unicorns that dominate the headlines."

It’s worth breaking down the list. Though I expect it will look different in a year, it’s

Researchers at the AI company Anthropic say they have made a fundamental breakthrough in our understanding of exactly how large language models, the type of AI responsible for the current boom, work. The breakthrough has important implications for how we may be able to make AI models safer, more secure, and more reliable in the future.

One of the problems with today’s powerful AI that is based around large language models (LLMs) is that the models are black boxes. We can know what prompts we feed them and what output they produce, but exactly how they arrive at any particular response is a mystery, even to the AI researchers who build them.

This inscrutability creates all kinds of issues. It makes it difficult to predict when a model is likely to “hallucinate,” or confidently spew erroneous information. We know these large AI models are susceptible to various jailbreaks where they can be tricked into jumping guardrails (the limits the AI model developers try to put around

Software major Microsoft announced on March 25 its six new Agentic Artificial intelligence (AI) agents designed to autonomously assist with critical areas such as phishing, data security, and identity management.

This is important as the company now processes 84 trillion signals daily, including 7,000 password attacks per second. Scaling cyber defenses through AI agents is now imperative to keep pace with this threat landscape.

"We are expanding Security Copilot with six security agents built by Microsoft and five security agents built by our partners—available for preview in April 2025. The relentless pace and complexity of cyberattacks have surpassed human capacity and establishing AI agents is a necessity for modern security," the company said in a release.
Microsoft launched its earlier version of Security Copilot a year ago to empower defenders to detect, investigate, and respond to security incidents swiftly and accurately.

Between January and December 2024,

REDMOND, SEATTLE: Australia can be a credible global force in the artificial intelligence boom by focusing on its strengths in developing applications and operating data centres, Microsoft chief executive Satya Nadella says.

Nadella, one of the world’s most powerful tech executives, said that China’s low cost artificial intelligence model DeepSeek had opened up the AI race, giving countries outside the US a chance to vie for leadership in aspects of AI technology.

“The biggest lesson of DeepSeek is let us not assume that the last big breakthrough in terms of model architecture has happened,” Nadella told The Australian at Microsoft’s headquarters in Seattle.

“We should not take too much for granted that whatever is the current AI regime is the only real regime going forward. Whether it’s in Australia or whether it is in any country – even in the global south – quite frankly, I think [anyone] has a shot at being able to innovate, sometimes on top of existing models

Snowflake Inc. is an American cloud-based data storage company. Headquartered in Bozeman, Montana, it operates a platform that allows for data analysis and simultaneous access of data sets with minimal latency. It operates on Amazon Web Services, Microsoft Azure, and Google Cloud Platform. As of November 2024, the company had 10,618 customers, including >800 members of the Forbes Global 2000, and processed 4.2 billion daily queries across its platform.

History

Snowflake Inc. was founded in July 2012 in San Mateo, California, by Benoît Dageville, Thierry Cruanes and Marcin Żukowski. Dageville and Cruanes previously worked as data architects at Oracle Corporation; Żukowski was a co-founder of Vectorwise. Mike Speiser, a venture capitalist at Sutter Hill Ventures, which provided early funding to the company, was the company's first CEO.

In June 2014, Bob Muglia, formerly of Microsoft, was named CEO. In October

Google CEO Sundar Pichai calls AI a “golden age of innovation,” emphasizing its transformative impact despite being in its early days. Speaking at the Paris AI Summit, he warned that failing to embrace AI is the biggest risk. As Google invests $75 billion in AI-driven growth, he outlined four key priorities: innovation, infrastructure, workforce readiness, and responsible development. While optimistic, he urged global leaders to address AI challenges like accuracy and the digital divide.

AI’s Potential and Challenges

Pichai emphasized that AI adoption must be guided by responsible policies to benefit everyone. He acknowledged concerns about misinformation, misuse, and inequality in access. He urged governments to develop strategies that balance innovation with regulation through investments in infrastructure and training.

World leaders and tech figures are meeting in Paris for a two-day artificial intelligence summit on Monday, aiming to discuss and determine the future direction of the rapidly developing industry.

Hosted by French President Emmanuel Macron, the AI Action Summit will welcome OpenAI CEO Sam Altman, US Vice President JD Vance and Chinese Vice Premier Zhang Guoqing.

France and Europe must seize the “opportunity” because AI "will enable us to live better, learn better, work better, care better and it’s up to us to put this artificial intelligence at the service of human beings,” he said.

The summit, which gathers major players such as Google, Microsoft and OpenAI, aims at fostering AI advances in sectors like health, education, environment and culture.

A global public-private partnership named “Current AI” is to be launched to support large-scale initiatives that serve the general interest.

The Paris summit “is the first time we’ll have had such a broad

U.S. officials are flaunting the idea of an outright ban on the sales of a popular Chinese-made internet router – a device that is sitting in many homes.

The investigation comes as concerns have mounted over cybersecurity risks connected with router manufacturer TP-Link Technology Co. The device is listed as a best seller on Amazon, with prices ranging from $50 to $100 for most models.

Established in China, but sourced from a California unit, the TP-Link router use has sparked several probes into possible Chinese interference through cyberattacks launched from the devices.

Several federal agencies, such as the Commerce, Defense, and Justice departments, are investigating the product and the company behind it. All are looking at if the devices pose a security risk and could suggest a ban as soon as 2025, company sources recently

China puts its communications network to the test
Chinese researchers have subjected data link hardware to extreme conditions to test its resilience, the South China Morning Post reported. These tests not only exceeded existing military standards, but reflect China’s commitment to the robustness of its communications systems in critical situations.

Evaluate the resilience of infrastructure
The simulations included scenarios such as nuclear explosions in the stratosphere, known as electromagnetic pulse (EMP) attacks. This approach aims to ensure that China's critical communications remain operational even in the event of a nuclear attack, strengthening its ability to respond in high-tension situations.

Quantum technology and reinforced hardware
It also promotes the development of quantum communication systems and reinforces military hardware, with the aim of strengthening its infrastructure against possible crises and guaranteeing

A ninth U.S. telecoms firm has been confirmed to have been hacked as part of a sprawling Chinese espionage campaign that gave officials in Beijing access to private texts and phone conversations of an unknown number of Americans, a top White House official said Friday.

Biden administration officials said this month that at least eight telecommunications companies, as well as dozens of nations, had been affected by the Chinese hacking blitz known as Salt Typhoon.

But Anne Neuberger, the deputy national security adviser for cyber and emerging technologies, told reporters Friday that a ninth victim had been identified after the administration released guidance to companies about how to hunt for Chinese culprits in their networks.

The update from Neuberger is the latest development in a massive hacking operation that has alarmed national