Soon-Next-Later is a structured futurology framework for categorising AI capability developments across three temporal horizons: Soon (0–5 years, capabilities already emerging in products), Next (5–10 years, capabilities requiring current research to mature), and Later (10+ years, speculative capabilities dependent on fundamental advances). The framework provides practitioners with a tractable planning scaffold that avoids both near-term over-hype and long-horizon dismissiveness.
Semantic Classification
Content
Soon, Next, and Later
- Digital Literacy, Data Privacy, and Algorithmic Bias (sorting the data)
- Understanding AI and digital technologies for information and service access.
- Ensuring privacy and mitigating biases in AI algorithms.
- 2-5 years of integration with business processes.
- Models start to distribute across cloud and devices to find their correct scale fit.
- Biome AI, AI Ethics & Safeguarding, and Ambient Education
- Integration of AI in biological lives and everyday environments.
- Developing robust ethical frameworks for AI.
- Creating ambient Education and AI environments facilitated by AI. Education is too slow and monolithic for this to happen sooner.
- 5-10 years.
- Fully Autonomous Agents and AI as lifelong companions
- AI operating without human supervision in various sectors. Value globally is arbitraged by these agents in near real-time. A renaissance of the idea of smart efficient markets?
- Global access to hyperlocal models which match context and personal choices, through personal devices which need not be cloud connected.
- AI understanding human emotions for psychological, contextual, and business support.
- No idea… I have no idea…
- Celebrating Human Diversity: Now to Later
- AI understanding and adapting to human experiences and identities.
- AI supported creativity: Soon to Next
- AI assisting and leading in art, music, and literature creation.
- Equity: Now to Next
- Democratizing resources and Education and AI globally.
- Self Guided Learning for Children: Next to Later
- Playful, memorable AI education experiences for children, globally.
- The age of the productive tinker: Later
- AI revolutionizing industries with specialized gadgets and applications.
- Resilience and collaborative management : Soon to Later
- AI optimizing resource consumption and enhancing waste management.
- AI’s role in climate change and wildlife conservation.
- Supporting our place: Later
- AI monitoring and managing physical health and environmental choices.
- Personal Health Management: Now to Next
- Personalized AI systems optimizing individual health. Things like [DermAssist
- Google Health](https://health.google/consumers/dermassist/) are just the start.
- Towards Conversational Diagnostic AI is scoring higher than clinicians already.
- Bill Gates views AI as a transformative tool in healthcare, particularly for enhancing access to education and mental health services.
- Lifetime Support Structures: Now to Later
- AI providing lifetime guidance and customized assistance.
- Effective collapse of integrated knowledge : Next to Later
- This is the 2016 Death of the Internet [Conpiracy Theory]([Dead Internet theory
- Wikipedia](https://en.wikipedia.org/wiki/Dead_Internet_theory)) but will likely happen in time
- Money will change: Next to Later
- Algorithms managing financial transactions and negotiations.
- Ubiquitous multi-modal UX: Next to Later
- Rise of integrated displays and interfaces for AI interaction.
- Intentional UX for accessing diverse networks of information: Later
Sectors
- Generative AI enhancing business productivity and efficiency through domain-specific models, optimizing workflows, and integrating multimodal applications.
- Advantage for employees who intersect with AI tooling, increasing engagement and productivity… but for the benefit of whom Employment Social Contract Under Automation
- Most time benefit from things “close to the metal” such as Dev Ops, coding, macros, light financial work, etc. Thing someone “looked up” then did.
- Predictive analytics and strategic insights transforming decision-making processes (white collar jobs).
- AI-driven project management which doesn’t sound like much but… Diagrams as Code
- Enhanced virtual Collaborative Systems Modality environments(think reduced travel for the classic complex visual tasks)
- Remote working; Metaverse and Telecollaboration (notes, minutes, knowledge management tools like this one). This includes radically improved document understanding
- Business to business mixed reality. I have been doing this since 1997. It’s always going to be big “SOON”, but the signs are improving and I can’t discount it as easily this time.
- Concentrate on the lowest hanging fruit, most people are not using this, you’ll get huge advantage. Protect your data, let the market develop. If you have a clear business case then do get a consultant and legal advice (£1000/hr). Make a risk matrix, use my GPT if you like.
- AI agents managing emails and consumer services, offering personalized and intuitive user experiences.
- [outfit anyone](Outfit Anyone (humanaigc.github.io))
- AI in personal finance, health monitoring, and personalized entertainment.
- AI enhancing environmental controls and home automation systems.
- AI streamlining healthcare administration and patient care, leading to efficient drug discovery and personalized medicine.
- AI’s expanded role in telemedicine and remote patient monitoring.
- AI crucial in cybersecurity, adapting to evolving threats and enhancing AI Governance Law and Privacy << this feels like it will be warfare
- Development of AI algorithms for adaptive threat response and robust Distributed Identity authentication processes.
- AI’s will demonstrate potential in personalised learning and administrative efficiency in educational settings, nobody will use it.
- Challenges and uneven adoption across institutions due to budget constraints, lack of expertise, and ethical concerns.
- Individuals who can will use major platforms like OpenAI for education, centralising innovation, and raising privacy and equity issues.
- Discussions will have to start on AI literacy, and collaborative approaches to ensure ethical and effective AI integration in education.
- Strategies for educators, technologists, and policymakers to foster an environment where AI benefits learning outcomes and is accessible to all.
- I think the Rabbit is something I would buy for kids?! (lol, that didn’t work out)
- AI spurring innovation in coding, product development, and creative fields like storytelling & games.
- Perhaps less so in image and video.
- AI increasingly integrated as a collaborative partner in content creation and user experience design.
- AI enhancing real-time data analysis and handling unstructured data for deeper insights.
- AI contributing to environmental and climate modeling for sustainable solutions.
- Addressing AI-related societal and ethical challenges, focusing on transparency, fairness, and accountability.
- Public discourse on AI’s implications on privacy, employment, and societal norms.
- AI hardware advancements leading to more energy-efficient and powerful processing capabilities.
- The beginning of AI Agents
- intentional UX.
- Having your personal agent, or it’s agents, do the online work for you,
- Bringing back distilled updates to a locally or securely hosted core agent.
- This will have staggering repercussions for the web as we know it.
- This is the old Death of the Internet conspiracy theory, but happening.
- Semantic and natural language programming
- In a future where intentional programming user experience (UX) evolves to its next stage, we could envision a scenario where multimodal and language models interface directly with bytecode, driven by semantic instructions from product owners. This approach would represent a significant leap from current programming paradigms, emphasizing a more intuitive, less syntax-heavy interaction with software development.
- 1. Semantic Instruction and Bytecode Manipulation:
- In this future, product owners or non-technical stakeholders could provide instructions in natural language or through other intuitive interfaces.
- These instructions would be semantically analyzed by advanced language models, capable of understanding the intent and context of the request.
- The language models would then translate these semantic instructions into bytecode – the lowest-level code executed by the computer’s processor.
- This process bypasses traditional programming languages, allowing for more direct and efficient creation or modification of software functionalities.
- 2. Multimodal Interfaces:
- Multimodal interfaces, incorporating voice, text, and possibly visual or gestural inputs, would make the process more accessible and intuitive.
- These interfaces would cater to a diverse range of users and preferences, allowing instructions to be given in various formats.
- 3. Enhanced Collaboration and Iteration:
- By enabling product owners to directly communicate their requirements to the software, the gap between idea conception and implementation narrows significantly.
- This direct communication loop would facilitate rapid iterations, as changes can be implemented and reviewed in real-time or near-real-time.
- 4. Automated Testing and Validation:
- The system would incorporate sophisticated automated testing mechanisms.
- As soon as the bytecode is generated or modified, a series of automated tests could run to validate the functionality, ensuring that the changes meet the specified requirements and do not introduce bugs or vulnerabilities.
- Product owners could also perform their own tests on the resultant functionality, using user-friendly testing tools integrated into this system.
- This is taken from Sam Hammond AI Policy Economist who I have discovered recently. All his stuff is summarised and linked here.