Author Archives: Originmeister

The artificial intelligence world has been rocked by recent claims from Chinese startup DeepSeek, which asserts it has developed an AI model rivalling those of tech giants like OpenAI and Google at a fraction of the cost. However, as the dust settles, scepticism is growing about the truth of these claims.

The DeepSeek Sensation

DeepSeek, a Hangzhou-based company backed by hedge fund High Flyer, made waves with the release of its DeepSeek-R1 model. The company claims that R1 matches or even surpasses the performance of OpenAI’s ChatGPT o1 on multiple key benchmarks. What’s more, DeepSeek asserts that it achieved this feat using only about 2,000 older generation Nvidia chips, at a cost of just $6 million in computing power. If true, this would mean that DeepSeek requires well under 10% of the processing power used by popular models from the likes of OpenAI, Google or Anthropic.

This announcement sent shockwaves through the tech industry. Nvidia, the leading manufacturer of AI chips, saw its stock plummet, losing nearly $600 billion in market value. The implications were clear: if DeepSeek’s claims were true, it could upend the entire AI industry, challenging the massive investments made by US companies.

Scepticism Emerges

As the initial shock subsides, many experts are starting to cast doubt on DeepSeek’s statements, the primary point of contention being the company’s claim about its limited use of computing resources.

Alexandr Wang, CEO of Scale AI, publicly challenged DeepSeek’s narrative. Speaking at Davos, Wang stated, “My understanding is DeepSeek has 50,000 H100s,” referring to Nvidia’s most advanced AI chips. This figure is dramatically higher than the 2,000 lower-power chips DeepSeek claims to have used.

Analysts have cast further doubt, pointing out the unlikelihood that a small startup could have discovered such enormous gains in efficiency that were meanwhile missed by very well-funded western research teams. Wall Street analysts Bernstein said “In short, we believe that 1) DeepSeek DID NOT “build OpenAI for $5M”; 2) the models look fantastic but we don’t think they are miracles”.

The Chip Question

Confusion is further fuelled by the context of US export controls on advanced semiconductors to China. These restrictions were specifically designed to limit China’s access to cutting-edge AI technology. If DeepSeek indeed has access to 50,000 H100 chips, it would not only contradict their public statements but also raise questions about the effectiveness of these export controls.

Transparency Issues

Adding to the scepticism is DeepSeek’s lack of transparency. While the company has open-sourced its model, allowing for some level of scrutiny, the details of its training process and resource usage remain murky. This opacity makes it challenging for independent researchers to verify DeepSeek’s claims.

The Broader Implications

If DeepSeek’s claims prove to be exaggerated or false, it could have significant repercussions. The company’s assertions have already influenced market dynamics and policy discussions; the market shock alone might be seen as a positive outcome by Chinese state propagandists. US President Donald Trump referred to DeepSeek’s rise as a “wake-up call” for the American tech industry.

Many experts argue that even if DeepSeek’s claims are inflated, they still highlight important trends in AI development. The focus on efficiency and resource optimization, driven partly by export controls, could lead to genuine innovations in the field that would be beneficial globally: lower costs would reduce barriers to entry, allowing smaller companies to develop their own models. The environmental benefits of reduced power consumption would also be very significant.

Looking Ahead

As the debate continues, the world eagerly awaits more evidence to support or refute DeepSeek’s claims. Regardless of the outcome, this episode underscores the intense competition in the AI sector and the geopolitical tensions surrounding technological advancement.

The true test will be whether DeepSeek can consistently deliver on its promises and compete with established players in the long run. As Sam Bresnick, a research fellow at Georgetown University’s Center for Security and Emerging Technology, noted, “The real test of their effectiveness will be whether U.S. firms are able to continue to outpace China in coming years”.

In the meantime, the DeepSeek story serves as a reminder of the need for transparency and verifiable claims in the fast-paced world of AI development. As the industry moves forward, the ability to balance innovation with credibility will be crucial for any company seeking to make its mark in this competitive landscape.

Both Google and Microsoft have recently positioned AI at the core of their commercial offerings. Microsoft 365 Copilot now offers integrated AI capabilities across their Microsoft 365 suite and Azure cloud services. Similarly, Google now bundles Gemini AI within its Workspace offerings, and provides a range of powerful AI services to its Cloud customers.

Both companies are at the forefront of addressing the ethical issues and legal risks associated with AI technology, and are taking proactive steps to reassure their corporate and public sector customers that AI can be implemented safely and responsibly. This article examines and compares the approaches taken by Google and Microsoft to mitigate potential problems and foster trust in AI adoption.

Ethical Frameworks and Principles

Google and Microsoft have each established comprehensive ethical frameworks based on the same six principles:

1. Fairness

2. Reliability and safety

3. Privacy and security

4. Inclusiveness

5. Transparency

6. Accountability

Governance and Implementation

Both companies have established robust governance structures to ensure the implementation of their ethical principles.

Google’s Approach

Google has created a federated, bottom-up approach combined with strong top-down support from company leadership. Over 200 Google full-time employees are dedicated to implementing ethical AI practises. Key components of their governance structure include:

1. AI Principles reviews conducted by dedicated teams

2. Consultations with internal experts for teams developing AI applications

3. Adversarial proactive fairness (ProFair) testing

4. Risk rating assessments focusing on potential impacts on people and society

Google has also invested heavily in internal education, with over 32,000 employees engaging in AI Principles training since 2019.

Microsoft’s Approach

Microsoft has implemented a multi-layered governance structure to operationalize its responsible AI practices:

1. Microsoft Board oversight through the Environmental, Social, and Public Policy Committee

2. Responsible AI Council co-led by senior executives

3. Office of Responsible AI (ORA) tasked with policy governance and implementation

4. AI Ethics and Effects in Engineering and Research (Aether) Committee for thought leadership

5. Engineering teams creating AI platforms, applications, and tools

Microsoft has also developed the Responsible AI Standard, which defines goals, requirements, and practices for all AI systems developed by the company.

Addressing Specific Concerns

Both Google and Microsoft are tackling key ethical and legal issues that concern their corporate customers.

Data Privacy and Security

Google emphasizes the importance of privacy and security in its AI systems, ensuring that they respect privacy and maintain the security of private and confidential information. The company has implemented strong safety measures at the AI model, platform, and application levels.

Microsoft has integrated privacy and security considerations into its Responsible AI Standard and requires compliance with existing privacy and security programs. The company also provides guidance on data privacy concerns related to its AI tools.

Intellectual Property and Copyright

Both companies are grappling with the complex legal landscape surrounding AI and copyright. Google is actively participating in discussions and legal proceedings related to the use of copyrighted material in AI training.

Microsoft has entered into agreements with publishers, such as Taylor & Francis, to improve the performance of its AI products while addressing copyright concerns. The company is also updating its contracts and terms of service to clarify the responsibilities of users regarding copyright and intellectual property.

Transparency and Explainability

Google has developed resources like the TensorFlow open-source toolkit to provide model transparency in a structured, accessible way. The company also emphasizes the importance of explainable AI in building trust with users and stakeholders.

Microsoft has incorporated transparency requirements into its Responsible AI Standard and provides documentation on its AI systems to help users understand how decisions are made. The company also conducts regular audits of its AI systems to ensure ongoing compliance with ethical guidelines.

Corporate Customer Reassurance

Both Google and Microsoft are taking steps to reassure their corporate customers that they can use AI safely and responsibly.

Google provides resources and best practices for responsible AI development, including fairness guidelines and technical references. The company also offers AI ethics training and workshops to help customers implement ethical AI practices within their organizations.

Microsoft has developed the AI Business School to help organizations understand the implications of AI and implement it responsibly. The company also provides transparency documents and resources to help customers navigate the ethical and legal landscape of AI adoption.

Conclusion

While the two tech giants have taken similar approaches to addressing the ethical issues and legal risks surrounding AI, there are some nuanced differences in their strategies. Google appears to place a stronger emphasis on internal education and bottom-up implementation of ethical principles, while Microsoft has developed a more structured, top-down governance model.

Both companies recognize the importance of transparency, accountability, and ongoing assessment in building trust with their corporate customers. By providing comprehensive frameworks, tools, and resources, Google and Microsoft are working to create an environment where organisations can leverage AI technologies while minimizing potential ethical and legal risks.

As the AI landscape continues to evolve, both companies will likely need to adapt their approaches to address new challenges and regulatory requirements. Public sector customers should remain engaged with these developments and work closely with their AI providers to ensure responsible and compliant implementation of AI technologies within their organisations.

The UK government has unveiled a comprehensive “AI Opportunities Action Plan” that aims to transform public services and boost economic growth through artificial intelligence. Prime Minister Sir Keir Starmer announced the initiative on 13 January, backed by £14 billion in private sector investment from tech companies including Vantage Data Centres, Nscale, and Kyndryl.

Key Components

Infrastructure Development The plan includes establishing AI Growth Zones across the country, with Culham, Oxfordshire designated as the first zone. A new state-of-the-art supercomputer will be developed to enhance the nation’s AI research capabilities.

Public Sector Integration A new digital centre within the Department for Science, Innovation and Technology (DSIT) will implement a ‘Scan, Pilot, Scale’ approach to AI adoption across public services. The government expects AI implementation to improve various aspects of public service delivery, from personalizing education to streamlining administrative tasks.

Economic Impact The International Monetary Fund estimates that widespread AI adoption could boost UK productivity by 1.5% annually, potentially generating £47 billion in economic benefits each year. The initiative is expected to create 13,250 new jobs.

The government will provide further details of the strategy in the upcoming Digital and Technology Sector Plan, with the Compute Strategy expected in Spring 2025.

Bramble Hub AI Partnerships

Numerous Bramble Hub partners provide AI skills and products that can be procured through us via public sector frameworks. These include companies like MyCardium AI, which applies AI in healthcare diagnostics, The Honest Data Company for predictive modeling and AI consulting, CloudSource, which is bringing Microsoft AI skills to the regulatory sector, and various other partners specializing in artificial intelligence solutions for public sector applications.