Humans: AI Legacy versus AI Generative AI AGI
In any technology innovative cycle there is also a legacy and the new players. As we move to the next wave, or should we say, iteration of Artificial Intelligence we have to understand the AI legacy of major global players and the new kids in block of Generative AI.
No new technology comes from the dark. There are always legacy and developments and when it comes to AI we are talking about decades of hard work of organisations, academic research and corporations building models and innovative cycles and no end followed by new more auspicious waves.
1. Legacy AI 2013 to 2022
From 2013 to 2022 the largest patent owners in machine learning and AI worldwide by number of active patent families were the following:
- Baidu
- Chinese Academy of Science
- Samsung
- Alibaba Group
- Tencent
- IBM
- Huawei
- State Grip Cord
- Ping An Insurance
- Microsoft
During 2013 to 2022dates IBM and Microsoft are the only major Western players leading this list. IBM ranked fifth with just under 9,500 active patent families. The statistic is based on data provided by PatentSight. According to the same source, in December 2022, Baidu was the largest owner of active machine learning and artificial intelligence (AI) patent families worldwide with 13,993 active patent families owned. In 2022, the company had claimed the leading position from Tencent, now ranked second with 13,187 active patent families owned.
One has to recall that IBM was the first big player in tech and in AI. And also the incredible number of IP. IBM has a total of 122110 patents globally, out of which 35080 have been granted. Of these 122110 patents, more than 59% patents are active. The United States of America is where IBM has filed the maximum number of patents, followed by China and Japan.
The period from 2013 to 2022 witnessed substantial advancements and shifts in the field of Artificial Intelligence (AI), marking significant milestones in its evolution. This era, often referred to as the period of “Legacy AI,” is characterised by the maturation of AI technologies, widespread adoption across various industries, and the emergence of ethical and governance discussions. Here’s an overview of key developments and trends during this period:
1.1. Advancements in Deep Learning:
- Breakthroughs in Neural Networks: This period saw significant improvements in deep learning algorithms, particularly in Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). These advancements led to major breakthroughs in computer vision, natural language processing (NLP), and other AI fields.
- AlphaGo’s Victory: In 2016, DeepMind’s AlphaGo defeated a world champion in the game of Go, a milestone demonstrating the potential of deep learning in solving complex problems.
1.2. Proliferation of AI in Industry:
- Business Adoption: Businesses across sectors started adopting AI for various applications, including customer service (AI chatbots), sales (predictive analytics), and operations (automation).
- Healthcare Innovations: AI began playing a crucial role in healthcare, from drug discovery to diagnostic tools like IBM’s Watson, which can assist in identifying treatment options for cancer patients.
1.3. Rise of AI Ethics and Governance:
- Awareness of AI Bias and Fairness: Issues related to AI bias, fairness, and transparency gained prominence, leading to increased research and discussions on ethical AI.
- Regulatory Initiatives: Governments and international bodies began developing frameworks and guidelines to govern AI usage, like the EU’s General Data Protection Regulation (GDPR) which includes provisions related to AI and data privacy.
1.4. Expansion of AI Tools and Platforms:
- Open Source and Cloud Platforms: Tools like TensorFlow and PyTorch democratised access to AI technologies, while cloud platforms like AWS, Google Cloud, and Azure made powerful computing resources accessible to a wider audience.
- AI-as-a-Service (AIaaS): Companies started offering AI services, allowing businesses to integrate AI capabilities without developing the technologies in-house.
1.5. Natural Language Processing (NLP) and Generative AI:
- Advancements in NLP: Progress in models like GPT (Generative Pre-trained Transformer) and BERT (Bidirectional Encoder Representations from Transformers) significantly improved language understanding and generation capabilities.
- Generative AI: Tools capable of creating content, whether text, images, or videos, began emerging, paving the way for creative and content generation applications.
1.6. Challenges and Controversies:
- Privacy Concerns: The use of AI in surveillance, data analytics, and personal assistants raised concerns about privacy and data security.
- Job Displacement Fears: The potential for AI and automation to displace jobs led to discussions about the future of work, re-skilling, and the social implications of advanced AI.
The period from 2013 to 2022 set the stage for the current state of AI, laying the technological foundation, raising awareness about ethical considerations, and integrating AI into the fabric of business and society. The legacy of this era is a robust, dynamic field that continues to evolve and shape various aspects of our world.
This is before Generative AI revolution. 2023 changed and is still revolutionising all things with Generative AI leapfrogging the AI landscape in a dangerous (even if exciting) path to AGI.
2. So where do we stay now when it comes to AI IP and patents?
I am trying to understand what will happen with the patents and specially with AI patents in a time of generative AI. As the innovation and disruption goes on, this new ecosystem is very different from anything in the past.
The velocity and the impact took the world by storm and at the moment Microsoft and openAI have a big percentage of the market followed by Google and Amazon.
3. Generative AI from 2023
In the bellow graphic and research from 2023 from iot-analytics.com we can see how Generative AI is taking the AI legacy world!
As one can see the Generative AI market share in 2023 was taken on storm by literally 4 players:
OpenAI 39%
Microsoft 30%
AWS 8%
Google 7%
Others 16%
So this is a new game as the other 16% including IBM that were the major legacy players will have to redesign a new strategy and adapt to this new world.
Moreover, one important, very important subject is that we are now in a time when it is no longer Humans that can create AI patents, patents can now be created by or with AI. In fact all creation of AI and tech now is done with some type of generative AI tool. So where do we stay when it comes to creating patents?
When it comes to AI-related intellectual property (IP) and the creation of patents, there are several international considerations to keep in mind. These considerations are shaped by the rapid evolution of AI technology, its cross-border nature, and the diverse legal frameworks across different countries. Here are the major international considerations:
3.1. Patentability Criteria:
- Innovation and Non-Obviousness: The AI invention must be new, involve an inventive step (non-obviousness), and be capable of industrial application. These criteria can vary significantly between jurisdictions.
- Subject Matter Eligibility: Some countries have restrictions on patenting algorithms or mathematical methods, which can affect AI patents. The interpretation of what constitutes patentable subject matter in the context of AI differs from one jurisdiction to another.
3.2. Ownership and Inventorship:
- Human vs. AI Inventorship: There’s an ongoing debate about whether AI systems can be named as inventors on patent applications. Currently, most jurisdictions require a human inventor, but this area is evolving.
- Employee and Contractor Inventions: The ownership of AI-created IP can be complex, especially when AI developers, data scientists, and other contributors are involved. Clear agreements and policies are essential.
3.3. Disclosure Requirements:
- Sufficient Disclosure: Patent applications generally require a detailed disclosure of the invention so that an expert in the field can replicate it. For AI, this can involve disclosing algorithms, training methods, datasets, etc.
- Trade Secrets vs. Patents: Some businesses may choose to protect their AI innovations as trade secrets, especially when the full disclosure required for patenting could reveal sensitive information.
3.4. Data Rights and Privacy:
- Use of Data: AI systems are often trained on large datasets. Ensuring that the use of this data complies with international data protection laws (like GDPR) is crucial.
- Data Ownership: The ownership of data used to train AI systems can impact the patentability and commercialisation of AI inventions.
3.5. International Harmonization and Treaties:
- Patent Cooperation Treaty (PCT): The PCT allows inventors to seek patent protection internationally for their inventions, simplifying the process of filing patents in multiple jurisdictions.
- Lack of Harmonisation: Despite treaties like the PCT, significant differences remain in how AI-related IP is treated across different jurisdictions, requiring a tailored approach for each region.
3.6. Enforcement Challenges:
- Cross-border Enforcement: Enforcing AI patents can be challenging, especially when infringement occurs across different jurisdictions with varied legal frameworks.
- Rapid Technological Evolution: The fast pace of AI development can make patents obsolete quickly, and businesses may need to continuously innovate to stay ahead.
3.7. Ethical and Social Considerations:
- Access to Technology: There’s a growing debate about the balance between protecting AI innovations and ensuring broad access to AI technology for societal benefit.
- Bias and Fairness: The way AI systems are designed and the data they are trained on can have significant social implications, influencing discussions about the ethical use and protection of AI technology.
Navigating the landscape of AI IP and patents requires a careful and informed approach, considering the legal, ethical, and practical implications across different international jurisdictions. Consulting with IP attorneys who specialize in AI and staying abreast of evolving regulations and norms is essential for businesses operating in this space.
In 2023, a court case of Thaler v. Vidal, the US Federal Circuit affirmed that only natural persons (i.e., human beings) can be named inventors on U.S. patents, and thereby have excluded artificial intelligence from being listed as an inventor per se. 43 F.4th 1207 (Fed. Cir. 2022).
source https://lnkd.in/eYSWpz6y.
4. Redefine AI and Human Inventions with AI
At the moment we have to redefine the work in AI, the definition of human inventions and human and machines creating inventions.
I would be able to affirm that we are now in a situation where around 99% of any invention done at the moment in the world is done, using some type of AI generative model. And inventions directed to Artificial Intelligence are being analysed by governments and organisations all over the world. That is the subject matter of AI related inventions that have been examined in the U.S. Patent and Trademark Office (“USPTO”) and issued as patents for many years already.
For example, the number of U.S. patent applications for AI-based chemical inventions more than doubled from 2009 to 2019, and issued U.S. patents in the same subject area more than tripled during the same time period. Currently, the situation takes new directions and AI-related inventions typically encompass advances in the AI architecture, computational techniques, hardware/material components, and specific uses of artificial intelligence. But the question now is that humans do it with AI and increasingly will be doing it.
With generative AI taken over, the importance is to manage the legacy AI research with all new development and make sure AI governance and AI security are managed to scale in a way to serve all Humans.
In the same context is important and interesting the road map of IBM for AI 2023 to 2030:
The road map of research from IBM highlights that large-scale self-supervised neural networks, i.e., foundation models, multiply the productivity and the multi-modal capabilities of AI. More general forms of AI emerge to support reasoning and common-sense knowledge
https://www.ibm.com/roadmaps/ai/
5. Ethical/ Trustworthy AI for Business
How do we define ethical/ Trustworthy AI for Business? What are the best practices when it comes to put a strong Ethical/ Trustworthy AI for Business. What multiple countries and governments advice and what are the best steps forward?
Creating a strong ethical and trustworthy AI framework for businesses involves adhering to principles that ensure AI systems are designed and operated in a way that is ethical, transparent, accountable, and respects user privacy and rights. Best practices in this area are informed by guidelines and frameworks provided by multiple countries and international organisations. Here are some key principles and steps forward:
5.1. Transparency and Explainability:
- AI systems should be transparent. Users should understand how and why a particular AI decision or output was generated.
- Adopt explainable AI (XAI) practices where the decisions made by AI systems can be understood and traced by humans.
5.2. Accountability and Oversight:
- Implement governance frameworks that ensure accountability for AI systems and their outcomes.
- Ensure that humans remain in the loop, especially for critical decision-making processes, to oversee AI operations.
5.3. Fairness and Non-discrimination:
- Test AI systems for bias and take steps to mitigate any unfair treatment of individuals or groups.
- Ensure that AI systems do not perpetuate, amplify, or lead to discriminatory outcomes.
5.4. Privacy and Data Governance:
- Adhere to data protection laws such as the GDPR in the EU, CCPA in California, or other local data protection regulations.
- Implement strong data governance practices to ensure the integrity and confidentiality of the data used and generated by AI systems.
5.5. Safety and Reliability:
- Design and deploy AI systems that are safe and reliable.
- Regularly test AI systems to ensure they function as intended and do not pose unforeseen risks.
5.6 Human-Centric and Social Well-being:
- Ensure that AI technologies respect human rights and values and contribute positively to societal well-being.
- Engage stakeholders including users, civil society, and domain experts in the design and deployment of AI systems.
5.7 Environmental and Societal Impact:
- Consider the environmental footprint of AI systems, especially the energy consumption of training large models.
- Evaluate the broader societal impacts of AI, including effects on employment and the distribution of economic benefits.
5.8. Steps Forward for Implementation:
- Establish Ethical Guidelines: Develop and adopt clear ethical guidelines for AI use within the organisation.
- Incorporate Multi-stakeholder Perspectives: Include diverse perspectives in the development and deployment of AI, considering the potential impact on various stakeholders.
- Continuous Monitoring and Evaluation: Regularly assess AI systems for compliance with ethical principles and legal requirements.
- Training and Awareness: Ensure that employees, especially those involved in AI development and deployment, are trained on ethical AI principles and practices.
- Regulatory Compliance: Stay informed about and comply with international, national, and local regulations governing AI.
- Engage with Industry Initiatives: Participate in industry initiatives and collaborations that aim to promote the ethical use of AI.
By adhering to these best practices and taking proactive steps, businesses can ensure that their use of AI aligns with ethical standards and societal values, fostering trust and ensuring the responsible and beneficial use of AI technologies.
6. Why is Governance critical for AI?
Governance is the cornerstone for advancing and scaling artificial intelligence.
Governance is critical for AI in business because it ensures that AI systems are used responsibly, ethically, and effectively, aligning with both business objectives and societal norms. Good governance helps mitigate risks, including legal and ethical violations, repetitional damage, and operational inefficiencies. Here’s why governance is essential for AI in business and how to create a robust governance framework, with reference to IBM’s governance principles as an example:
6.1. Importance of Governance for AI in Business:
- Ensuring Ethical Use: Governance frameworks guide businesses in the ethical use of AI, ensuring that AI systems respect user privacy, operate transparently, and do not perpetuate bias or discrimination.
- Compliance with Regulations: With the evolving landscape of AI regulations (e.g., GDPR in Europe), governance ensures that AI practices comply with legal standards, avoiding penalties and legal issues.
- Risk Management: Proper governance identifies and mitigates risks associated with AI, including data breaches, unintended consequences of AI decisions, and potential misuse of AI technologies.
- Building Trust: Transparent and responsible use of AI fosters trust among customers, partners, and regulators, which is crucial for the long-term success of AI initiatives.
- Enhancing Performance and Accountability: Governance ensures that AI systems perform as intended and that there is accountability for their outcomes, which is essential for maintaining operational integrity and achieving business objectives.
6.2. Creating Good AI Governance Principles:
- Define Clear Objectives and Scope: Establish what the organisation aims to achieve with AI and the boundaries within which AI systems should operate.
- Establish Ethical Guidelines: Develop a set of ethical principles that AI systems and their developers must adhere to, covering fairness, transparency, accountability, and respect for user privacy.
- Implement Oversight Mechanisms: Set up governance bodies or committees responsible for overseeing AI initiatives, ensuring they align with the organisation’s ethical principles and business objectives.
- Ensure Transparency and Explainability: Make AI systems and their decisions understandable and transparent to users and stakeholders, fostering trust and enabling informed decision-making.
- Promote Data Governance: Implement robust data management practices to ensure the quality, integrity, and security of the data used in AI systems.
- Regular Monitoring and Compliance Checks: Continuously monitor AI systems for compliance with governance principles, legal standards, and ethical guidelines, and conduct regular audits.
- Stakeholder Engagement: Involve various stakeholders, including customers, employees, and regulators, in the governance process to ensure diverse perspectives and needs are considered.
6.3. Case study of IBM’s AI Governance Principles:
IBM is known for its commitment to ethical AI and has established a set of governance principles, which include:
- Transparency and Explainability: IBM insists on clear communication about how and when AI is being used, and decisions made by AI systems should be explainable.
- Accountability: IBM holds that while AI can augment human intelligence, the final decision-making authority should rest with humans, ensuring accountability.
- Fairness and Bias Mitigation: IBM actively works to develop and deploy AI systems free of bias, ensuring that they treat all users fairly.
- Privacy and Security: IBM commits to the highest standards of data privacy and security, ensuring that user data is protected and AI systems are secure from threats.
By incorporating these principles and practices, businesses can establish a robust governance framework for AI, ensuring that their AI initiatives are ethical, compliant, and aligned with both their business objectives and societal values.
IBM emphasises the significance of an ethical, AI-centred approach to governance. Their framework involves a wide range of stakeholders including AI developers, users, policymakers, and ethicists, ensuring that AI systems align with societal values. This comprehensive involvement is crucial for developing and using AI-related systems responsibly.
IBM’s AI Governance Consulting services underscore the transformative potential of AI in business. They offer expertise in helping enterprises leverage AI to drive business transformation and harness the value from AI-induced disruptions. This approach emphasises the strategic integration of AI governance into business practices, ensuring that AI initiatives are aligned with business objectives and deliver substantial value.
Additionally, IBM’s AI Academy on AI Governance provides resources for setting up responsible AI workflows and outlines the overall process of AI activities in an organisation. This guidance is geared towards ensuring that organisations’ AI initiatives result in trusted outcomes and explainable results, highlighting the importance of transparency and accountability in AI operations.
These principles and resources from IBM offer a robust framework for businesses looking to incorporate AI into their operations ethically and effectively, ensuring that AI governance is an integral part of their organisational strategy.
As a new and important strategy that expands the scope of the challenges and opportunities with AI IBM Consulting is expanding its strategic expertise to help clients / partners adopt responsible AI practices, encompassing automated model governance and broader organizational governance. This includes addressing AI ethics, organizational culture, accountability, training, regulatory compliance, risk management, and cybersecurity threats.
The watsonx.governance offering is part of the IBM watsonx AI and data platform, alongside other products like AI assistants and data storage solutions, designed to assist enterprises in scaling and accelerating their AI initiatives. Additionally, IBM is offering intellectual property protection for its IBM-developed watsonx models.
The broader watsonx portfolio allows IBM aims to enable businesses to innovate with AI while maintaining transparency, accountability, and control over their AI initiatives.
7. AI and business partners ecosystems
Why AI needs business partners ecosystems that comprehend the academia, governmental organisations, corporations and partner network ecosystems?
Artificial intelligence needs strong partner ecosystems. AI technology is complex, multifaceted, and has far-reaching implications across various sectors. To effectively develop, deploy, and govern AI, it’s crucial to have a diverse ecosystem that includes academia, governmental organisations, corporations, and partner networks. Here’s why this comprehensive ecosystem is essential:
- Innovation and Research (Academia): Universities and research institutions are at the forefront of AI research. They play a crucial role in advancing the underlying technologies, exploring new applications, and providing a deep understanding of theoretical aspects. Collaborations with academia can fuel innovation, provide access to top talent, and help in keeping abreast of the latest developments in AI.
- Regulation and Standards (Governmental Organizations): AI has significant societal, ethical, and legal implications. Governmental organisations are essential for setting regulations and standards that ensure the responsible development and use of AI. They can also support AI initiatives through funding, infrastructure, and policy-making, ensuring that AI serves the public interest.
- Application and Scaling (Corporations): Corporations can apply AI to solve real-world problems and scale solutions across markets. They have the resources, infrastructure, and channels necessary to bring AI innovations to market. Corporations can also provide practical insights into how AI can drive business value, which is crucial for making AI solutions viable and sustainable.
- Collaboration and Synergy (Partner Network Ecosystems): Partner networks, including technology providers, service firms, and industry consortia, facilitate collaboration and sharing of best practices. They can offer specialised expertise, access to unique datasets, and the ability to co-create solutions. These networks are essential for fostering a collaborative environment where different entities can work together to advance AI technology and its applications.
7.1. Benefits of a Comprehensive AI Ecosystem:
- Holistic Development: This ecosystem ensures that AI is developed with a comprehensive understanding of its technological, ethical, social, and commercial aspects.
- Responsible AI: Collaboration between these stakeholders can promote the development of responsible AI that aligns with ethical standards, regulatory requirements, and societal values.
- Accelerated Innovation: An ecosystem approach facilitates the exchange of ideas and resources, accelerating the pace of AI innovation and its application to real-world problems.
- Risk Mitigation: A diverse ecosystem can provide a robust framework for identifying, assessing, and mitigating risks associated with AI, from data privacy issues to the potential impacts on employment.
- Global Standards and Policies: Engaging a broad range of stakeholders can help in developing global standards and policies for AI, promoting consistency and interoperability across different regions and industries.
By fostering partnerships across academia, government, corporations, and partner networks, the AI ecosystem can drive innovation, ensure responsible use, and maximise the societal and economic benefits of AI technologies.
8. Humans Legacy and Humanity 2.0
Artificial Intelligence
Better than humans at one specific task
Most current Al applications
- Deep Blue
- Siri
- Alexa
- DeepL Translator
- Self driving cars
- Healthcare AI
- Financial processing intelligent data
- Sophia the Robot
- AiDa the robot
AGI Artificial General Intelligence
Capable of every task like humans
Envisioned but not yet realised mostly present in fiction
- R2D2, C-3PO (Star Wars)
- Samantha (Her)
- Ava (Ex Machina)
- Winston (Origin)
Humans AI Action Plan
Algorithms are and will increasingly hire and firing us now.
As we move into AGI the highest risk category is now defined by the number of computer transactions needed to train the machine, known as “floating point operations per second” (Flops).
Can we create a Human-Centered AI? The global ecosystem of AI needs to develop with a human-centric approach respecting fundamental rights, human values, building trust, building consciousness of how we can get the best out of this AI revolution that is happening before our eyes. Respecting the Humans: AI Legacy in a time of Gernative AI moving to AGI.
Each of us, our businesses, our governments and the world have to work to set in place a real blueprint and clear path that includes education + regulation for AI. Our patents, our inventions, our intelligence are being reengineered. Our human present and future — our very own intelligence for the future of us and our children is a digital world driven by AI. We need to have an action plan guided by an ecosystem that fosters the development and evolution of this technology in a human-centric ecosystem and direction.
