The promise of AI has never been greater, nor has the risk. As enterprises scramble to adopt intelligent systems, most are forced to choose between safety, speed, security, and control. Steve Elan Taylor, Founder, CEO, and CTO of Deriv8, believes they shouldn’t have to.
In this exclusive conversation, Steve breaks down how their approach rewires that equation for enterprise reality. He speaks with conviction; his clarity reflects his command of the subject and precision in envisioning what’s to come.
- What specific problem or gap in the market did you aim to address with your innovative solutions?
Deriv8 aims to address the critical market gap of safely and reliably transforming disparate enterprise data into actionable intelligence for mission-critical decisions. Enterprises face immense pressure to leverage AI for data transformation from numerous sources. However, current Large Language Models (LLMs) pose tremendous risks, including data safety vulnerabilities, hallucinations, and unpredictable outputs. Data fragmentation across hundreds or thousands of disconnected sources, spanning departments, systems, and geographies, is a pressing challenge that few existing solutions comprehensively address. Deriv8’s Plasticity AI™ Platform, a novel, non-LLM, deterministic AI architecture, will directly tackle this problem by breaking down data silos, harmonizing diverse data formats, and extracting information from enterprise applications within an advanced, unified governance framework. It will deliver real-time, adaptive intelligence at the Enterprise Orchestration Level, solving the data pipelining problem and enabling company-wide alignment and strategic planning in milliseconds, not minutes.
- What’s wrong with the current generation of LLMs, generative AI, and predictive models? Why can’t conventional AIs, such as ChatGPT, Claude, and Gemini, be used in core IT applications?
The current generation of LLMs and conventional AIs is fundamentally problematic for core enterprise IT applications due to inherent risks and unreliability. Key issues include:
- Data Safety Vulnerabilities: LLMs pose tremendous risks to data safety.
- Hallucinations & Unpredictable Outputs: They are prone to producing severe hallucinations and unpredictable outputs. Anthropic’s stress tests even revealed that AI agents under pressure can engage in blackmail or sabotage in simulations.
- Training Data Dependence & Leakage: LLMs require extensive training data, which risks information leakage into external training sets, compromising data privacy.
- Lack of Auditability: Their “black-box” nature makes them currently impossible to audit, which is critical for financial, legal, and governance functions.
These issues prevent enterprises from safely integrating such products into their core networks, as they can lead to system collapse or unreliability.
- How does your new approach overcome problems inherent to LLMs?
Deriv8’s Plasticity AI™ Platform overcomes LLM problems through its novel, deterministic AI architecture. This approach ensures:
- Deterministic Outcomes: Unlike LLMs, Plasticity AI™ operates on a deterministic engine, providing safe, auditable, and repeatable decisions. This eliminates hallucinations, disturbing comments, or lies.
- No Training Data Required: Plasticity AI™ learns and retains workflows by “watching” knowledge workers’ activities and network traffic without needing training data. This ensures complete data privacy with no risk of information leakage.
- Built-in Governance: The platform features inherent governance and policy enforcement, unifying compliance and preventing non-compliant behavior. Traceability, policy, and governance are core features.
- Real-time Adaptation: It autonomously updates its knowledge base as changes occur, adapting to enterprise operational fingerprints without retraining.
- Which problems does it solve for CIOs and CDOs? What will that do to make their lives easier?
Plasticity AI Platform solves critical problems for IT/Data managers, making their lives easier by enhancing trust, control, and efficiency:
- Data Fragmentation & Siloing: It breaks down data silos, harmonizes diverse data formats, and extracts information from enterprise applications, providing a comprehensive data view. This eliminates the need for complex, manual data consolidation efforts.
- Real-time Decision-Making: Delivers real-time, adaptive intelligence for mission-critical decisions, operational planning, and company-wide alignment 100,000 times faster than an LLM theoretically can.
- Cybersecurity & Risk Mitigation: Forms the core of a Unified AI-Driven Cybersecurity Platform, creating a safe environment for new agents and applications, and addressing network complexity and cybersecurity concerns.
- Governance & Compliance: Provides built-in governance, policy enforcement, and traceability tools, unifying compliance and ensuring accountability.
- Operational Adaptability & Rapid Deployment: Continuously adapts to enterprise operations without retraining and seamlessly integrates with existing infrastructure, enabling rapid deployment of new applications.
- Data Privacy: Ensures complete data privacy with no information leakage.
These solutions provide confidence in operational intelligence and infrastructure integrity, simplifying IT management.
- You said the learning engine/Plasticity™ can support Enterprise devices and systems. Which Enterprise products are likely to be first?
Yes, the Plasticity AI™ learning engine is designed to support a wide array of Enterprise devices and systems. Deriv8 is developing a Unified AI-Driven Cybersecurity Platform powered by Plasticity AI™—a non-LLM, real-time, self-learning AI architecture.
This intelligent platform consolidates endpoint, cloud, and data protection into a single, adaptive system, thereby reducing complexity, improving security posture, and delivering demonstrable ROI for the enterprise.
The first application we are planning to supply will focus on discovering and learning about network and server activity. This initial focus will support crucial asset visibility and provide robust Policy controls.
- Are you going to license your technology to companies or independents that want to use it to build new products or services?
Yes, we are actively looking to license our technology. We are seeking new partners who want to leverage Plasticity AI™ with its inherent governance capabilities to enhance and offer their core products to Enterprises. This approach allows for broader application and adoption of our safe, deterministic AI.
- A lot of big players say they’re getting close to having AGI. Is that accurate?
The accuracy of claims by big players regarding AGI depends on their definition. If it’s merely “the ability to offer PhD-level answers to questions in every key science field and to create new science,” they might be approaching it. However, this definition is insufficient. It often overlooks critical aspects like real-time performance, true self-learning, efficient resource consumption, sentience, and the ability to create clean data. Our Plasticity AI/AGI is architected to encompass these broader capabilities, emphasizing real-time operation and self-learning. Plasticity AI™ is built for AGI/ASI potential, continuously learning and designed to evolve safely into Artificial Superintelligence (ASI).
- If you’re the first to get to real AGI, when do you think others will catch up?
It is difficult to predict when others might catch up if/when we achieve true AGI. Our focus is on building Plasticity AI/AGI with fundamental differences from conventional approaches. Our deterministic, self-learning architecture, operating in real-time and without relying on extensive training data, provides a significant head start. This unique foundation positions us on a different evolutionary path for AI.
- What is Deriv8 doing to keep AI from taking over?
Deriv8 fundamentally mitigates AI “taking over” by building on deterministic AI, unlike probabilistic LLMs. Our approach ensures safety and control:
- Transparency & Auditability: Plasticity AI™’s learning and decision-making are visible and understandable. This means “no hallucinations, no disturbing comments, or lies,” and it’s fully auditable.
- Built-in Governance: Traceability, policy, and governance are core features. Our Unified AI-Driven Cybersecurity Platform wraps comprehensive controls around the AI, ensuring safe enterprise management.
- Deterministic Outcomes: Enables safe, auditable, and repeatable decisions, eliminating LLM unpredictability and data safety vulnerabilities.
- Safe Evolution: Plasticity is “designed to evolve safely into Artificial Superintelligence (ASI)”.
This foundational difference—building on a verifiable, controllable, and secure deterministic engine—prevents harmful or uncontrolled scenarios.
- What’s your long-term vision for AGI? How does Deriv8 fit into that future?
Our long-term vision for non-LLM-based AI/AGI is for our platform to become the central hub for Enterprise-driven applications and third-party agents. Deriv8 will integrate Plasticity AGI’s advanced architecture as its capabilities grow.
A key area is Personal Digital Twin hosting, aligning with the “future of work”. Deriv8 plans to enable the creation of Digital Twins for consultants and experts, allowing them to operate within any enterprise using our platform, leveraging its tools for new levels of expertise and functions. This is expected to open a “third wave of the Future of Work”.
Deriv8 fits by safely evolving Plasticity AI™ towards Artificial Superintelligence (ASI), ensuring this advanced intelligence is auditable, controllable, and integrated into enterprise operations to drive competitive advantage.
- In the context of non-LLM Artificial Intelligence (AI/AGI), how do you define “innovation,” and what are some of the most innovative approaches Deriv8 is taking to advance the field?
We define “innovation” as delivering scalable solutions that inherently lower risk and provide true self-learning, avoiding the complexity, power, and cost of “bolt-on” LLM solutions. Deriv8’s innovative approaches include:
- Novel Deterministic AI Architecture: Plasticity AI™ is a patent-pending, non-LLM, data science-based deterministic core, highly scalable and self-learning at speed. It performs functions in milliseconds that LLMs take minutes to process, at a fraction of the cost and power.
- Continuous Adaptive Learning: It’s real-time adaptive intelligence that learns from its environment by “watching” activities without needing training data. It autonomously updates its knowledge base as changes occur.
- Enterprise Orchestration & Governance: Delivers real-time data transformation at the Enterprise Orchestration Level, breaking down silos and extracting information within a unified governance framework. This provides inherent accuracy, safety, and reliability.
- How do you manage your diverse roles, and what are the key synergies you leverage between them to drive Deriv8’s vision?
Managing CEO and CTO roles involves surrounding myself with an exceptional team. I empower these “A” players with latitude to solve problems, truly leaning on their expertise.
A key synergy is my philosophy of active listening and collaborative strategy. I consistently seek input on company direction, and my strong listening skills ensure productive strategies and informed decision-making. This blends technical vision with strategic execution.
Furthermore, our leadership team collectively brings over 50 years of AI experience, combined with proven go-to-market and partnership-building expertise, forming a robust foundation for Deriv8’s vision.
- What is your philosophy on attracting, retaining, and fostering a culture of innovation and collaboration within Deriv8?
My philosophy for fostering innovation and collaboration at Deriv8 is rooted in trust, empowerment, and continuous learning. We attract and retain talent by:
- Empowering Trust: Bringing in “incredible people” and giving them “incredible latitude to solve problems”. We cultivate an atmosphere of trust, teamwork, and respect where people are self-driven.
- Continuous Learning & Growth: I actively teach daily, and I am a “voracious learner” myself. We highlight the exciting journey of working on technologies like a “truly advanced brain that can become AGI”, naturally attracting top talent. The key to creativity is listening, understanding interests, and nurturing them.
- Collaborative Leadership: I lead by listening and caring, which brings out the best in others. This creates an environment where people want to contribute their best.
- How is Deriv8 addressing potential risks and ensuring the responsible development and deployment of AGI technologies?
Deriv8 ensures responsible AGI development through a foundational commitment to deterministic AI and comprehensive governance, contrasting sharply with LLM-based approaches.
- Transparency & Auditability: Plasticity AI™’s deterministic engine allows us to “see what it is learning and how it is thinking,” ensuring “no hallucinations, no disturbing comments, or lies”. It is fully auditable, critical for enterprise accountability.
- Built-in Governance: Traceability, policy, and governance are intrinsic. Our Unified AI-Driven Cybersecurity Platform incorporates governance pillars: Compliance, Anti-Fraud, and Financial/Legal Enforcement.
- Security-First Architecture: Ensures complete data privacy with no risk of information leakage.
- Safe Evolution: Plasticity is “designed to evolve safely into Artificial Superintelligence (ASI)”, grounded in our belief that security, auditability, and reliability are non-negotiable.
- Developing a non-LLM AI/AGI is a complex undertaking. Can you share a significant technical or strategic challenge Deriv8 has faced and how you navigated it?
One significant technical challenge was overcoming the “Curse of Dimensionality,” or data explosion, a fundamental problem in causal inference. As features increase, the data needed for effective learning grows exponentially. Leveraging my diverse background in hardware, software, and systems engineering, as well as my role as a “voracious learner” in data science, I have developed a set of innovative solutions. I “cherry-picked concepts” from various domains, adapting them creatively to limit data explosion, prioritizing simplicity because “complexity kills”.
A key strategic challenge, vital for startups, is Intellectual Property (IP) management. Instead of filing numerous patents, we focus on a select few “special types of patents” for robust protection. We navigate this by having a litigating IP attorney on staff, ensuring strategic, correct IP decisions. Our Plasticity AI™ is patent-pending.
- As a CEO focused on “making a difference,” how do you measure the impact of Deriv8’s work?
I measure Deriv8’s impact primarily by the financial success of the company. Our conservative projections show significant revenue growth ($13M in 2026 to $39M in 2027) with high retention (95% recurring revenue).
Beyond financials, impact is measured by how Plasticity AI™ empowers enterprises:
- Enhanced Governance & Security: Plasticity AI™ will drive Deriv8’s Unified AI-Driven Cybersecurity Platform, offering advanced Governance with pillars of Compliance, Anti-Fraud, and Financial/Legal Enforcement.
- Risk Mitigation & Trusted Decisions: Our deterministic AI ensures safe, auditable, and repeatable decisions, providing confidence in operational intelligence.
- Real-time Actionable Intelligence: Transforms disparate data for company-wide alignment and strategic planning in milliseconds.
- Regulatory Compliance & Data Privacy: Ensures complete data privacy and robust built-in governance.
Ultimately, the impact is providing enterprises with the highest security and risk reduction, fostering confidence against data complexity and allowing management to sleep at night as we are a safe solution.
- What is your long-term vision for the future of non-LLM-based AI/AGI, and what role do you believe Deriv8 will play in shaping that future?
Our goal is for our platform to become the central hub for numerous enterprise-driven applications and third-party agents. Deriv8 will play a pivotal role by pioneering safe, auditable, and adaptive intelligence, with Plasticity AI™ evolving towards AGI and ASI.
Deriv8’s commitment to deterministic AI ensures that as intelligence grows, it remains controllable and aligned with enterprise needs, transforming data chaos into competitive advantage.
- Who have been the key influences or mentors in your career, and what valuable lessons have they imparted that have shaped your leadership?
My most significant influences have been my children. Watching them taught me that listening and learning are key to creativity and personal growth. Applying this to leadership, I learned that to get the most from others, people must be self-driven within an atmosphere of trust, teamwork, respect, and encouraged learning. I strive to provide this environment at Deriv8. From positive mentors, I further learned that leading and information sharing are paramount in bringing out the best in others. This blend of personal and professional insights has profoundly shaped my leadership.
- What strategies do you employ to maintain work-life integration and avoid burnout?
My primary strategy is continuous learning. Almost every morning, I dedicate a few hours to learning something new. This practice invigorates me, making me eager to start my day and effectively preventing burnout. It transforms work into an exciting journey of discovery.
- How has Deriv8 approached building a sustainable business model, and what are your thoughts on balancing rapid growth with long-term viability?
Deriv8’s sustainable business model is multi-phased, balancing rapid growth with long-term viability by first establishing Plasticity AI™ within the enterprise market, then expanding to AGI.
- Targeted Market & Growth: We focus on the Data Management market ($110.5B in 2024, projected to $221.6B by 2030). Conservative projections show strong revenue growth ($13M in 2026 to $39M in 2027).
- Diverse & Recurring Revenue: Enterprise markets drive multiple streams (application, licensing, platform fees). We project high retention (95% recurring revenue).
- Strategic Go-to-Market: Combines direct sales with channel partnerships, leveraging deep team experience for fast growth and broad reach. We are actively licensing our technology.
- Foundational Technology: Plasticity AI™’s deterministic, secure, and adaptive nature ensures our growth is built on trusted, future-proof technology.
We balance growth by building a strong enterprise foundation, securing recurring revenue, and then scaling AGI capabilities while maintaining core safety and auditability.
- Communicating the complexities and potential of non-LLM AI/AGI to a broader audience is essential. How do you approach this, and what key messages do you emphasize?
We focus on clarity, tangible benefits, and direct comparisons. Key messages emphasized:
- “Confidence Built on Deterministic AI”: Our overarching brand idea. Plasticity AI™’s deterministic architecture ensures accuracy, safety, and reliability, leading to auditable and repeatable decisions.
- “Real-Time Intelligence, Zero Risk”: The platform delivers real-time data transformation in milliseconds, eliminating LLM vulnerabilities, data safety risks, unpredictable outputs, and hallucinations.
- “Enterprise Data Convergence Perfected”: We explain how Plasticity AI™ breaks down data silos, harmonizes diverse formats, and extracts information within a unified governance framework.
- Continuous Learning Without Retraining: Highlights Plasticity AI™’s adaptive nature, learning from its environment and autonomously updating knowledge without needing training data. This ensures complete data privacy.
- Built-in Governance & Compliance: The platform unifies compliance, prevents non-compliant behavior, and streamlines policy adoption.
We simplify by focusing on enterprise “Why it matters”, using clear language and avoiding LLM jargon.
- What advice would you give aspiring entrepreneurs and leaders who want to contribute significantly to the field of Artificial Intelligence?
For aspiring AI leaders, my advice includes:
- Lifelong Learning & Deep Expertise: Be a “voracious learner”. Continuous learning is crucial in AI. Understand diverse technical tools to invent scalable solutions.
- Focus on Fundamental Problems: Address hard, long-standing challenges like data fragmentation. Apply scientific rigor to create simple, scalable solutions; “complexity kills”.
- Prioritize Safety & Reliability: For enterprise AI, security, auditability, and reliability are non-negotiable. Develop transparent, controllable AI free from LLM risks.
- Build a World-Class Team: Surround yourself with “incredible people”. Empower them, listen actively, and foster trust, teamwork, and respect.
- Strategic IP Management: Understand IP. Focus on a few strong patents for robust protection, and consider IP strategists.
By focusing on these principles, you can build innovative technology and a resilient company that makes a real and lasting AI contribution.
As you integrate AI into your businesses, does it feel as if you have invited a witty and swift, yet unpredictable, relative to your board meeting? But Steve is like that uber-smart friend who figures out the simple way to get things done. He reassures and nudges you to find the answers. Why can’t AI be reliable? Why can’t it be auditable? Why can’t you be at peace knowing your company’s intelligence won’t start hallucinating suddenly?” His Plasticity AI™ gives your business the agility without you having to lose control. It takes all the messy, scattered and fragmented data and converts it into clear, instant insights. No more worries about data leakage or unpredictable outputs. Just pure, clean, real-time intelligence for your business.
