Intelligent or not, AI is incredibly useful and powerful
AI lacks genuine understanding, relying on data patterns and statistical analyses. It excels in processing and synthesising information, aiding research, creativity, and problem-solving. While not truly intelligent, AI tools enhance human capabilities and open new avenues for innovation and learning.
A new form of intelligence
The perception of AI as truly "intelligent" or "artificial" can be quite misleading. At its core, AI is built on extensive datasets gathered by corporations. Instead of having genuine understanding or consciousness, these systems are adept at recognizing patterns and conducting statistical analyses. They process huge volumes of data to produce outputs that resemble human responses. The impression of intelligence comes from their ability to offer a range of answers, especially when they revise and refine responses to meet certain expectations. However, this behaviour doesn’t reflect real reasoning; it's simply a result of their programming to improve outputs based on feedback. Essentially, AI serves as a powerful tool for data processing, much like a sophisticated search engine, but it still relies on human intelligence to interpret and validate its findings.
Nonetheless, AI systems like ChatGPT bring significant benefits despite these limitations. They excel at rapidly processing and synthesizing vast amounts of information, making them valuable assets for research, creative brainstorming, and problem-solving across various domains. By providing coherent and contextually relevant responses, they support users in overcoming challenges, streamlining decision-making processes, and enhancing overall productivity. In this way, while they do not possess human-like understanding, these AI tools effectively augment human capabilities and open up new avenues for innovation and learning.
Here is my take of AI tools that I have explored
Google Bard
Who Owns: Google
Market Share: Emerging (exact figures undisclosed)
Number of Users: Not publicly disclosed (early adoption phase)
Strengths:
Leverages vast search data and integrates seamlessly with Google's ecosystem.
Offers real-time information and continuous updates.
Weaknesses:
Early-stage product may sometimes produce inaccuracies.
Faces challenges in building long-term user trust.
Opportunities:
Can further integrate with a wide array of Google services and applications.
Potential to expand real-time, data-driven capabilities.
Threats:
Intense competition from other advanced AI models.
Regulatory and privacy concerns may impact adoption.
Microsoft Bing Chat
Who Owns: Microsoft
Market Share: Growing within Microsoft’s ecosystem (no precise public share)
Number of Users: Estimated tens of millions (integrated with Bing’s large user base)
Strengths:
Combines GPT-4 technology with robust integration into Bing and the Windows ecosystem.
Backed by Microsoft’s enterprise-grade infrastructure.
Weaknesses:
Mixed feedback on response consistency and coherence.
Heavily reliant on external model architectures for performance.
Opportunities:
Opportunity to leverage enterprise solutions and productivity tools.
Potential to refine and personalize search and conversational experiences.
Threats:
Strong competition from both Google Bard and ChatGPT.
Ongoing challenges in maintaining user trust and reliability.
Anthropic Claude
Who Owns: Anthropic
Market Share: Niche, with a smaller market footprint
Number of Users: Limited; estimated in the thousands to tens of thousands (pilot/limited access)
Strengths:
Emphasises safety, transparency, and ethical AI usage.
Produces contextually sensitive, user-aligned responses.
Weaknesses:
Limited language support and integration with mainstream consumer platforms.
Opportunities:
Can capture niche markets prioritizing ethical and safe AI.
Opportunity to form strategic partnerships in regulated industries.
Threats:
Faces stiff competition from larger, well-funded tech companies.
Market adoption hurdles if performance scaling is slow.
Meta LLaMA
Who Owns: Meta
Market Share: Not directly measured (open-source model)
Number of Users: Not directly tracked; widely used in research and custom implementations
Strengths:
Open-source model offering high customizability and adaptability.
Strong support from the research and academic communities.
Weaknesses:
Less refined for consumer applications; requires technical expertise.
Limited commercial polish compared to proprietary models.
Opportunities:
Drives innovation in academic research and niche commercial projects.
Community-driven enhancements can lead to diverse applications.
Threats:
Risk of fragmentation within the open-source ecosystem.
Vulnerable to inconsistent quality control and misuse.
Gemini
Who Owns: Google
Market Share: Not established (pre-release stage)
Number of Users: Not applicable; pending public release
Strengths:
Next-generation AI with potential for advanced reasoning and multimodal capabilities.
Backed by Google’s deep research resources and robust infrastructure.
Weaknesses:
New entrant with limited public exposure and real-world testing.
Early deployment may face reliability and scalability challenges.
Opportunities:
Opportunity to integrate deeply with Google’s expansive ecosystem.
Potential to set new standards in AI innovation and user interaction.
Threats:
High market expectations could lead to critical scrutiny.
Intense competition and regulatory challenges may impact growth.
Perplexity AI
Who Owns: Perplexity AI
Market Share: Niche with a modest user base
Number of Users: Not publicly disclosed; likely in the lower tens of thousands to low hundreds of thousands
Strengths:
Provides transparent, source-backed answers with a user-friendly interface.
Merges conversational capabilities with real-time search data effectively.
Weaknesses:
May not match the conversational depth and nuance of more established models.
Occasionally over-relies on search results, limiting creative responses.
Opportunities:
Growing market demand for integrated, verifiable AI responses.
Potential to enhance features through strategic partnerships and technology improvements.
Threats:
Rapid advancements in AI may outpace its current development.
Competing against larger, more resource-rich AI platforms.
GitHub Copilot
Who Owns: Microsoft (GitHub)
Market Share: Significant within the developer community (millions of active users)
Number of Users: Approximately 10 million active users (across various integrations)
Strengths:
Specialises in code generation and intelligent code suggestions.
Deep integration with popular IDEs like Visual Studio Code.
Powered by OpenAI Codex and supported by a vast developer community.
Weaknesses:
Limited to coding tasks rather than general conversation.
Can occasionally suggest inaccurate or insecure code, necessitating human oversight.
Opportunities:
Opportunity to expand integration across more development platforms and programming languages.
Rising demand for developer productivity tools can drive further adoption.
Threats:
Competition from other coding assistants (e.g., Tabnine, Amazon CodeWhisperer) and emerging general-purpose models with coding capabilities.
Legal and licensing challenges regarding training data and generated code.
Microsoft 365 Copilot
Who Owns: Microsoft
Market Share: Potentially large, leveraging Microsoft’s vast Office user base (precise share pending wider release)
Number of Users: Not yet widely available; anticipated to reach millions of enterprise users upon full release
Strengths:
Deeply integrated with the Microsoft Office suite (Word, Excel, PowerPoint, Outlook, Teams) to enhance productivity.
Leverages context-aware AI to assist with drafting, analysis, and data insights.
Weaknesses:
Primarily designed for the Microsoft ecosystem, limiting versatility outside Office applications.
Concerns over data privacy and the need for user adaptation.
Opportunities:
Opportunity to revolutionize office productivity and become an industry standard in enterprise environments.
Further integration with other Microsoft products and services could expand its reach.
Threats:
Competition from alternative productivity tools and similar AI integrations in other ecosystems (e.g., Google Workspace).
Possible resistance from users hesitant to adopt AI-driven workflows.
Roger Hunt
nem Partner
April 2025