Connect with us

Business

OpenAI rediscovers an open AI mission, with new reasoning models

Published

on


It was a long wait since the GPT-2 in 2019, but OpenAI is now releasing its newest open-weight large language models (LLMs). They’ve been dubbed GPT-OSS, the current lot consisting of gpt-oss-120b and gpt-oss-20b, dubbed “reasoning models” with OpenAI claiming these models outperform similarly sized open models on reasoning tasks. The importance of this brings OpenAI back, in a way, to their original mission of building AI systems that benefit all of humanity. Over the years, the artificial intelligence (AI) company has faced criticism of distraction towards that stated mission, as competition escalated rapidly.

OpenAI has not released open source models. (Official Image)

“Releasing gpt-oss-120b and gpt-oss-20b marks a significant step forward for open-weight models. At their size, these models deliver meaningful advancements in both reasoning capabilities and safety. Open models complement our hosted models, giving developers a wider range of tools to accelerate leading edge research, foster innovation and enable safer, more transparent AI development across a wide range of use cases,” the company says, in a statement.

Also read:Fidji Simo, OpenAI’s new CEO, insists AI can put power in the hands of people

Two questions that need to be answered before we get into the specifics of gpt-oss-120b and gpt-oss-20b models. First, what are open weight LLMs and are they different from LLMs you regularly use? And secondly, what are reasoning models? The former is best defined as a large language model that is released by a company publicly, in its entirety, which means all the actual model weights (read this as parameters, defined by billion or “b” in model names) and any user can download these models completely on their own hardware.

In comparison, the most popular LLMs that you may have used, including OpenAI’s own GPT models as well as the likes of Google Gemini 2.5 and Anthropic’s Claude Sonnet 4, are closed models — that means they are accessible through an application layer while the model weights are not in the public domain. At the same time, Meta’s Llama models, as well as certain models by Mistral, have followed the open weight methodology, in recent times.

Open weight AI models are not to be confused with open source models however, the fine difference being that the latter models such as the DeepSeek R1 also make training code, datasets, and linked documentation available publicly — open weight models don’t. Having the training code and data sets allows a user or developer to retrain an open-source model from scratch, often for customised usage scenarios. That flexibility isn’t there for open weight models, though accessible in their entirety. OpenAI has not released open source models.

To the second question, reasoning models slightly differ from a few other LLMs in the sense that they are specifically designed to spend more time “thinking through” complex problems before generating their final response. They are expected to use extended reasoning processes to work through multi-step problems.

Back to gpt-oss-120b and gpt-oss-20b, and the primary difference is in the number of parameters each has. Parameters are essentially like the strength of synapses in a human brain, which determines how strongly different “neurons” influence each other, before providing an answer for a query. In OpenAI’s naming scheme this time, there is a slight confusion — the gpt-oss-120b is a 117 billion parameter model, while the smaller gpt-oss-20b has 21 billion parameters.

OpenAI’s benchmark scores do peg the gpt-oss-120b and gpt-oss-20b close to the o3 and o4-mini models in most tests. Take for instance the MMLU benchmark, which consists of questions across academic disciplines — the gpt-oss-120b returned 90% accuracy while gpt-oss-20b clocked 85.3% accuracy; in comparison, the o3 (93.4%), o4-mini (93%) and o3-mini (87%) bookend the new open weight models.

Just in case you are wondering about the memory requirements for downloading and running these open weight models on your computing device, OpenAI confirms that the gpt-oss-120B model will need 80GB of memory on the system, while gpt-oss-20b requires at least 16GB. They say that Microsoft is also bringing GPU-optimised versions of the gpt-oss-20b model to Windows devices.



Source link

Business

Chennai’s BCS Launches Agentic isAI, a No-Code, Self-Orchestrating AI Built for Business Automation

Published

on


  • Agentic isAI is a no-code, autonomous AI platform that learns from behavior and automates enterprise operations without human prompts
  • Business Core Solutions is a Chennai-based tech company delivering enterprise-grade automation platforms trusted by global industry leaders

 

The Chennai-based technology company Business Core Solutions (BCS) has launched Agentic isAI, a path-breaking, autonomous AI platform that reimagines enterprise automation. This first-of-its-kind solution empowers businesses to proactively manage operations, reduce downtime, and automate critical workflows, all without the need for prompts, coding, or training data.

 

Chennai’s BCS Launches Agentic isAI, a No-Code, Self-Orchestrating AI Built for Business Automation

 

True to its name, which is inspired by the Tamil word “Isai” – meaning music, Agentic isAI has the ability to bring harmony and intelligence to complex enterprise systems. Unlike traditional AI systems that require extensive training data or manual prompts, Agentic isAI operates with true autonomy – it can observe system behaviour, detect anomalies, and initiate responses without human intervention. The solution is designed as a no-code platform – it learns from real-time patterns instead of relying on pre-trained models, making it faster to deploy and easier to adapt across industries.

 

Built for enterprise environments, it seamlessly integrates with platforms like SAP, Azure, Salesforce, Oracle, and more. Already in production with leading global clients, it is helping reduce downtime, prevent job failures, optimise cloud usage, and automate critical business operations at scale.

 

BCS is a Chennai-based enterprise technology company specialising in intelligent automation platforms for global businesses. Its flagship solutions, including Symphony for process orchestration, Anugal for identity governance, and DQView for data quality management, are trusted by Fortune 100 companies and industry leaders across the US, Europe, Middle East, and Asia-Pacific. With the launch of Agentic isAI, BCS continues to push the boundaries of scalable, AI-driven enterprise automation.

 

In his comments, Mr. Prakash Palani, Founder, Business Core Solutions, said, “We’re proud and excited to launch Agentic isAI, a product that reflects years of deep enterprise insight and cutting-edge innovation. What makes Agentic isAI truly path-breaking is its ability to act autonomously, without prompts, training data, or code, and still deliver reliable, enterprise-grade automation. We believe this platform has the power to fundamentally transform how businesses operate, making them more responsive, efficient, and resilient in a fast-changing world.”

 

He added: “Innovations like this are often expected to emerge from Silicon Valley or other global tech hubs. But Agentic isAI was imagined, engineered, and brought to life right here in Tamil Nadu. It proves that world-class enterprise technology can be built anywhere, as long as there is intent, talent, and vision. This launch isn’t just a milestone for us. It’s a moment of pride for the entire region.”

 

Beyond Agentic isAI, BCS offers a suite of powerful enterprise platforms designed to address core operational challenges. Symphony is an AI-powered orchestration platform that streamlines IT and business processes, enabling seamless automation across complex systems. Anugal focuses on identity and access governance, helping organizations ensure compliance, manage risk, and enforce security with precision. Meanwhile, DQView is a modern data quality platform that brings visibility, validation, and trust to enterprise data landscapes. These solutions reflect BCS’s commitment to building deeply integrated, scalable technologies that drive measurable impact for global businesses.

 

At the heart of BCS is a belief that technology and social responsibility can grow together. The company has consistently invested in inclusive hiring, with over half its workforce comprising individuals from underrepresented backgrounds, including first-generation graduates, rural youth, women returning to work, and persons with hearing and speech impairments. Through initiatives like the BCS Academy, which trains and places students, and HERizon, which supports women re-entering the workforce, BCS has created not just jobs, but opportunities for transformation. Its social initiatives also extend to improving public education, installing clean water systems, and nurturing local talent, making it a company where business success and human impact go hand in hand.



Source link

Continue Reading

Business

EasyJet captain suspended after getting ‘drunk and naked’ in hotel

Published

on


EasyJet has suspended one of its captains after he was reportedly seen roaming a luxury hotel drunk and naked.

The unnamed pilot was witnessed walking through common areas of a five-star resort in Cape Verde without any clothes on in the early hours of the morning on 5 August, after an extended drinking session in a bar, according to the Sun.

He was due to operate a return flight to Gatwick more than 36 hours later, but was grounded after the budget airline received complaints about the incident and a replacement pilot found.

An EasyJet spokesman told the BBC the pilot now faces an investigation and that the safety of passengers and crew was its “highest priority”.

The captain arrived at the Melia Dunas Beach Resort and Spa in the West African island nation on 4 August and proceeded to begin drinking, the Sun reports.

At around 02:30 local time (04:30 BST) the following morning, hotel guests reportedly saw him strip off and wander into the reception, before moving on to the gym and spa, according to the newspaper.

“The pilot did not have a stitch on and reeked of alcohol,” an anonymous source inside the airline was quoted by the paper as saying.

“Anyone who saw the pilot cavorting naked in the early hours on the day before a flight would not dream of getting on a plane with him at the controls.”

He was scheduled to helm the 2,332-nautical-mile (4,318km) trip back to Gatwick on the afternoon of 6 August, but was removed from the flight.

An EasyJet spokesman said: “As soon as we were made aware, the pilot was immediately stood down from duty, in line with our procedures, pending an investigation.

“The safety of our passengers and crew is EasyJet’s highest priority.”

The airline’s code of business ethics states that staff must behave “with integrity when dealing with our people, our customers, our partners and the communities within which we operate”.



Source link

Continue Reading

Business

KT reports record operating profit on robust AI business in Q2

Published

on


South Korea’s major mobile carrier KT Corp., which has picked artificial intelligence as its next growth engine, posted a record-high quarterly operating profit exceeding 1 trillion won ($719 million) in the second quarter, thanks to a double-digit growth in its AI business.

The company announced in a regulatory filing on Monday that its consolidated operating income came to 1.01 trillion won for the April-June period, more than doubling from the same period last year and breaching the 1 trillion won mark for the first time since its inception.

Revenue climbed 13.5% to 7.43 trillion won, while net profit surged 78.6% to 733.3 billion won.

Profit was partly driven by a one-off gain of 390 billion won from selling the company’s real estate assets.

But KT also attributed brisk sales in the business-to-business unit, especially AI services, to the stellar result.

Sales from its AI and information technology business jumped 13.8% on-year to 317.6 billion won in the quarter, whereas its mainstay wireless service added 1.6% to 1.7 trillion won. Broadband revenue also inched up 2.1%.

(Graphics by Daeun Lee) 

KT expects momentum to continue in the second half with the launch of new business-to-business products, including a large language model (LLM) and advanced cloud-computing services.

It also projects stronger subscriber growth in mobile after some SK Telecom Co. users switched carriers to KT following an April data breach that affected all of SK’s 25 million customers.

ACCELERATED TRANSITION INTO AICT COMPANY

The upbeat result in the AI business is expected to accelerate KT’s push to transform into a so-called AICT company, a plan unveiled by KT Chief Executive Kim Young-shub at this year’s MWC Barcelona in March.

AICT service is aimed at bolstering traditional telecommunications services with IT and AI technologies.

KT is Korea’s only company that has formed AI partnerships with both Microsoft Corp. and Palantir Technologies Inc.

It also signed a strategic partnership with Amazon Web Services (AWS) to bolster customized mobile and generative AI services for corporate customers in February last year.

It plans to introduce its LLM under development with Microsoft soon, as well as encrypted cloud services using confidential-computing technology, which processes encrypted data only after verifying a trusted execution environment. 

KT CEO Kim Young-shub (left) and Microsoft CEO Satya Nadella shake hands after signing a strategic partnership at Microsoft’s headquarters in Redmond, Washington on June 3, 2024 (Courtesy of KT) 

“In the second half, we will release our encrypted cloud developed with Microsoft and unveil a full AI lineup, including K-ChatGPT, a LLM model,” said Jang Min, KT’s chief financial officer. “The move will help KT cement its position as Korea’s leading AI company.”

COLLABORATION WITH US BIG TECH COMPANIES 

KT’s collaboration with Palantir in AI is also expected to gain further traction.

In March, it became the first Korean company to forge an exclusive strategic partnership with the US tech company.  

KT will incorporate Palantir’s AI and big data platforms with its own cloud and network infrastructure to help AI transformation, or AX, in both public and private businesses.

“We are now the most trusted AX partner in the financial industry,” said a KT official.

The utilization rate at a data center operated by KT Cloud, KT’s cloud service offering subsidiary, has already surpassed 90%, underscoring demand. 

After failing to join the state-led sovereign AI initiative, KT will focus more on the private sector through its collaboration with big tech companies, the company said.

Last week, the Korean government selected five local technology firms, including LG, Naver and SK Telecom, to spearhead the country’s flagship sovereign AI initiative, as Seoul moves to build large-scale AI models independent of US tech giants such as OpenAI, the operator of ChatGPT.

KT joined the race but dropped out of the bid in the second round of evaluation.

Write to Ji-Hee Choi at mymasaki@hankyung.com
Sookyung Seo edited this article.



Source link

Continue Reading

Trending