Alphabet’s Google has revealed two innovative AI systems, AlphaProof and AlphaGeometry 2, which demonstrate significant advancements in solving complex mathematical problems. These systems tackled abstract math more effectively than previous AI models, showcasing enhanced reasoning capabilities.
DeepMind, Google’s AI unit, reported that these models managed to solve four out of six questions at the 2024 International Math Olympiad. AlphaProof, which integrates the Gemini language model with the AlphaZero system, solved three problems, including the most challenging one, while AlphaGeometry 2 solved another.
These achievements mark the best performance by an AI system in the competition to date, with some problems solved in minutes and others taking up to three days. Meanwhile, Microsoft-backed OpenAI is developing a similar project known as ‘Strawberry,’ raising concerns among its staff about its potential impact on humanity.
Singapore-headquartered AI cloud provider Sustainable Metal Cloud (SMC) is set to expand globally, driven by fast-growing demand for its energy-saving technology. CEO and co-founder Tim Rosenfield announced plans to extend operations to EMEA (Europe, Middle East, and Africa) and North America in response to client demand. Currently, SMC operates “sustainable AI factories” in Australia and Singapore, with new launches planned in India and Thailand.
Partnering with AI chip giant Nvidia, SMC uses over 1,200 of Nvidia’s high-end H100 AI chips in Singapore to run open-source models like Meta’s Llama 2. Unlike most data centres that rely on air cooling technology, SMC employs immersion cooling, submerging Dell servers fitted with Nvidia GPUs in a synthetic oil called polyalphaolefin. The following method reduces energy consumption by up to 50% compared to traditional air cooling.
The International Energy Agency (IEA) anticipates a tenfold increase in AI demand compared to 2023, with global data centre electricity consumption expected to exceed 1,000 terawatt-hours by 2026. Sustainable Metal Cloud is currently raising $400 million in equity and $550 million in debt to support its expansion, according to sources. That move aligns with the increasing environmental concerns impacting Singapore’s data centre growth and highlights the importance of sustainable technology in meeting future energy demands.
Researchers at the Paul Scherrer Institute (PSI) and the Massachusetts Institute of Technology (MIT) have developed an AI system to improve the categorisation of breast cancer. The new technology, led by G.V. Shivashankar from PSI and Caroline Uhler from MIT, aims to provide a reliable and cost-effective method for predicting the progression of ductal carcinoma in situ (DCIS) to invasive ductal carcinoma (IDC).
DCIS, a precursor of breast cancer in the milk ducts, accounts for about 25% of breast cancer diagnoses. It can develop into a threatening invasive form in 30 to 50% of cases. The AI system, trained on tissue samples stained with DAPI dye, analyses chromatin images to identify patterns matching those identified by human pathologists. This approach leverages AI’s potential, as highlighted by research in Lancet Digital Health showing AI outperforming radiologists in breast cancer detection.
The researchers believe this AI-based tumour classification method has significant potential, though further studies are necessary to ensure its reliability and safety. The US Department of Defense (DoD) has been using AI to detect cancer since 2020, showcasing the growing role of AI in medical diagnostics. The new system developed by PSI and MIT could lead to more accurate predictions and better treatment decisions for patients.
Seomjae, a Seoul-based education solutions developer, is set to launch its AI-powered mathematics learning program at the Consumer Electronics Show in Las Vegas next January. The program uses an AI Retrieval-Augmented Generation model, developed over two years by a team of 40 mathematicians and AI developers. It features over 120,000 math problems and 30,000 lectures, offering personalised education tracks for each student.
Beta testing will begin on July 29, involving 50 students from Seoul, Ulsan, and Boston. The feedback will help enhance the technology and its feasibility. The innovative system, called Transforming Educational Content to AI, extracts and analyses information from lectures and problem solutions to provide core content.
Seomjae is also expanding its business portfolio to include an essay-writing educational program through partnerships in the US and Vietnam. The company will participate in Dubai’s Gulf Information Technology Exhibition this October, showcasing its new educational technologies.
A company official expressed excitement about starting beta testing and integrating diverse feedback to improve the program. The goal is to refine the AI system and ensure its effectiveness for students worldwide.
Grundon Waste Management is investing £750,000 in AI technology over three years to enhance driver safety. The company will implement Samsara’s Connected Operations Platform across its fleet of over 300 vehicles, following successful trials at two depots. The trials showed a 71% reduction in detected events and increased fuel efficiency, encouraging optimal driving habits.
Grundon expects to save around £220,000 annually in fuel costs once the technology is fully deployed. Chris Double, Regional Operations Manager, noted positive feedback from drivers during the trials. The AI Dash Cams, which provide instant feedback and acknowledge good performance, have been well-received.
The technology includes Dual-Facing AI Dash Cams and other cameras that monitor driver activity and connect to existing 360-degree cameras. Drivers can also use the Samsara App to track their behaviour through a points-based system. The system aims to improve safe driving habits and encourage good behaviour.
Philip van der Wilt, SVP and General Manager EMEA at Samsara, highlighted the measurable impact of the technology during the trials. He looks forward to a long-term partnership with Grundon to continue driving innovation and safety in their operations.
A recent survey conducted by Nikkei Research for Reuters shows a significant divide among Japanese companies regarding AI adoption. Of the 250 firms that responded, 24% have integrated AI into their operations, while 35% plan to do so. However, 41% have no plans to implement the technology, highlighting varying levels of enthusiasm for AI across corporate Japan. Key motivations for adopting AI include addressing workforce shortages, reducing labour costs, and accelerating research and development.
Despite Japan’s initially laid back approach to AI regulation, several obstacles hinder AI adoption, including employee anxiety over potential job losses, lack of technological expertise, high capital expenditure, and concerns about AI’s reliability. Cybersecurity also emerged as a critical issue, with 15% of respondents experiencing cyberattacks in the past year and 9% reporting attacks on their business partners. To enhance cybersecurity, 47% of firms outsource their defence, while 38% rely on in-house specialists.
The survey also touched on the controversial topic of Japan’s surname law, which mandates that spouses use the same surname. Half of the respondents support changing the law, a move prompted by the Keidanren business lobby’s recent appeal to the government. Supporters argue that the current system undermines individual dignity and freedom, particularly for women, while opponents believe separate surnames could weaken family bonds. Only 14% of firms believe the change would boost employee morale, and 10% think it would aid hiring efforts, with the majority expecting no significant impact on business.
Recent research from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has uncovered significant insights into the capabilities of large language models (LLMs). The study found that while LLMs excel in familiar scenarios, they struggle with novel tasks, raising questions about their true reasoning abilities versus reliance on memorisation.
The researchers compared LLMs’ performance on common tasks to hypothetical scenarios that deviated from their training data. For instance, models like GPT-4 showed proficiency in arithmetic using base-10 but faltered with other number bases, indicating a lack of generalisable addition skills. The pattern was consistent across various tasks, including spatial reasoning and chess, where models performed no better than random guessing in unfamiliar settings.
Lead author Zhaofeng Wu emphasised the importance of these findings, noting that as AI becomes more integrated into society, it must handle diverse scenarios reliably. The study’s insights aim to inform the development of more adaptable and robust future LLMs. The team plans to expand their research to include more complex and varied tasks, further exploring AI’s limitations and improving interpretability.
Supported by the MIT–IBM Watson AI Lab, the MIT Quest for Intelligence, and the National Science Foundation, the study was presented at the North American Chapter of the Association for Computational Linguistics (NAACL).
Elon Musk’s grand vision of establishing a human colony on Mars is rapidly taking shape at SpaceX, where intensive planning efforts are underway alongside rocket development. Musk, driven by a lifelong fascination with Mars, has directed SpaceX teams to design everything from dome habitats to spacesuits capable of withstanding Mars’s harsh conditions. His ambitious timeline now targets having a million people living on Mars within the next 20 years, a drastic acceleration from earlier projections.
Musk remains undeterred despite the immense challenges—such as freezing temperatures, dust storms, and the need for artificial atmospheres. SpaceX’s Starship rocket, designed to transport humans to Mars, is central to these plans, with recent successful test flights marking crucial milestones. The company envisions Starship as a transportation vessel and a potential living space equipped with amenities like living quarters and recreational facilities, crucial for long-term habitation on the red planet.
Asimov’s Foundation series was part of the inspiration for making life/consciousness multiplanetary.
Being multiplanetary greatly extends the probable lifespan of civilization.
While Musk’s vision has sparked both admiration and scepticism, particularly given the complexities of Martian colonisation, SpaceX is forging ahead with concrete steps. Internal discussions include considerations on bioengineering, sustainable living through greenhouses, and even potential food sources like plant-based alternatives from Impossible Foods. Despite operational challenges and controversies, SpaceX continues to attract dedicated employees who share Musk’s belief in creating a multi-planetary civilisation.
Musk’s determination to secure humanity’s future on Mars remains resolute, setting SpaceX on a course that challenges conventional timelines and expectations in space exploration. While NASA projects a much later timeframe for human missions to Mars, Musk’s aggressive pursuit of his Martian dream underscores his relentless drive to push the boundaries of what’s possible in space travel and colonisation.
OpenAI has launched a five-tier system to measure its progress towards developing AI that can surpass human performance. The new classification aims to provide clearer insights into the company’s approach to AI safety and future goals. The system, unveiled to employees during an all-hands meeting, outlines stages from conversational AI to advanced AI that are capable of running an entire organisation.
Currently, OpenAI is at the first level but is approaching the second stage, called ‘Reasoners.’ That level represents AI systems that can perform basic problem-solving tasks comparable to a human with a doctorate but without additional tools. During the meeting, leadership showcased a research project involving the GPT-4 model, demonstrating new capabilities that exhibit human-like reasoning.
The five-tier framework is still a work in progress, with plans to gather feedback from employees, investors, and the board. OpenAI’s ultimate goal is to create artificial general intelligence (AGI), which involves developing AI that outperforms humans in most tasks. CEO Sam Altman remains optimistic that AGI could be achieved within this decade.
The Nigerian Government has announced the development of a locally-made blockchain called ‘Nigerium’, designed to secure national data and enhance cybersecurity. The National Information Technology Development Agency (NITDA) is leading this initiative to address concerns about reliance on foreign blockchain technologies, such as Ethereum, which may not align with Nigeria’s interests.
NITDA Director General Kashifu Abdullahi introduced the ‘Nigerium’ project during a visit from the University of Hertfordshire Law School delegation in Abuja. He highlighted the need for a blockchain under Nigeria’s control to maintain data sovereignty and position the country as a leader in the competitive global tech landscape. The project, proposed by the University of Hertfordshire, aims to create a blockchain tailored to Nigeria’s unique requirements and regulatory framework.
The indigenous blockchain offers several advantages, including enhanced security, data control, and economic growth. By managing its own blockchain, Nigeria can safeguard sensitive information, improve cyber defence capabilities, and promote trusted transactions within its digital economy. The collaboration between the private and public sectors is crucial for the success of ‘Nigerium’, marking a significant step towards technological autonomy.
If successful, ‘Nigerium’ could place Nigeria at the forefront of blockchain technology in Africa, ensuring a secure and prosperous digital future. This initiative represents a strategic move towards maintaining data sovereignty and fostering innovation, positioning Nigeria to better control its technological destiny.
