AI & Web3: Understanding the Technologies Driving Future Innovation

The Spaces convened a practical, two-hour-plus webinar hosted by Abisola, featuring Hamid (smart contract/blockchain developer), Lamidi aka Lamcy (Web3 security researcher), Anderson Victor (software/machine learning engineer), and Timileni (AI engineer). The first half unpacked today’s Web3: why many entered early, how 2026 has shifted from hype (memes/airdrops) to real applications (tokenized value, privacy with zero-knowledge), and how decentralization, smart contracts, and user-owned data underpin the stack. Lamcy detailed the security researcher’s workflow (business logic comprehension, tooling like Foundry/Echidna/Medusa, CEI pattern, reentrancy, EVM atomicity/finality, formal verification) and shared how AI accelerated his ZK learning to find unique bugs. Hamid outlined blockchain developer responsibilities (modeling value flows, contract management, a security-first mindset) and beginner mistakes (learning in isolation vs. collaborating in community). The second half clarified AI vs ML, emphasized that ML is a subset powering AI systems, and stressed software engineering for model deployment. Anderson and Timileni offered career roadmaps (Python, data science, ML, NLP, deep learning, AI engineering), execution habits (avoid tutorial hell, build in public, document, ask the right questions), and portfolio strategies. Q&A covered smart contracts vs blockchains, model selection fit, and program logistics (recordings, track switching, application deadlines).

Webinar Notes — Web3 and AI Careers: Fundamentals, Roles, and How to Start

Session context and speakers

• Host: Abisola (webinar series for a boot camp cohort; emphasis on community, learning pace, and upcoming selection timeline and replays)
• Speakers:
  • Hamid — Smart Contract/Blockchain Developer and Tutor
  • Anderson Victor (aka “Dre”) — Software Engineer and Machine Learning Engineer
  • Lamcy (introduced as “alamendi”) — Web3 Security Researcher (smart contracts, protocols, ZK cryptography)
  • Timilani (referred to in the session as “Timmy Lane/Similian”) — AI Engineer (built AI systems across domains)

Web3

Why Web3: personal motivations and the state of the ecosystem

• Hamid: Entered Web3 because it was new, less saturated, and technically intriguing; stayed because of conviction in decentralization (finance, contracts) and the maturation of real use cases.
  • Then vs now: 2015–2017 era focused on adoption and hype (memes, NFTs, airdrops). By 2026, the noise is fading and meaningful applications are emerging (e.g., tokenizing real-world value/rewards, privacy via zero-knowledge). The space is evolving in the right direction with sustainable, practical use cases.

What Web3 and blockchain mean (for newcomers)

• Web3: The internet paradigm that leverages blockchain to shift power from centralized custodians to users.
• Blockchain: Decentralized, transparent, and tamper-resistant infrastructure where no single entity monopolizes data/control. It’s designed for trust-minimized coordination and verifiable fairness.
• Smart contracts: Code-as-agreement; terms are enforced by execution, not by centralized intermediaries. This underpins trustless applications (e.g., DeFi, DApps) and replaces reliance on institutional trust with cryptographic guarantees.

Careers in Web3 (beyond coding)

• It’s a broad ecosystem. Any profession can find a path provided you understand the fundamentals:
  • Technical: Smart contract engineering, protocol engineering, security research/auditing, cryptography (including ZK), infrastructure.
  • Non-technical/Hybrid: Technical writing, legal/compliance, product, community, marketing/growth, moderation, design.
• Core cross-cutting requirement: foundational literacy in blockchain concepts (what blockchains are, how smart contracts work, differences across chains and use cases). With that, you can transfer existing skills into Web3 contexts.

Web3 Security Research: what the work entails (Lamcy)

• Focus: Find bugs in smart contracts, protocols, and systems; increasingly includes ZK cryptography.
• Motivation: High frequency and impact of hacks; security is the “cybersecurity of blockchain.”
• Day-to-day/process:
  • Fundamentals first: understand how the target chain (e.g., EVM) works; grasp business logic, contract entry points, state transitions, money flow.
  • Read artifacts: project whitepapers, audit reports, post-mortems. Learn known attack vectors.
  • Core EVM attack primitives/patterns: reentrancy, Checks-Effects-Interactions (CEI), cross-contract/nested call pitfalls, state desynchronization, economic/logic flaws.
  • Tooling: Foundry for development/testing; fuzzers (e.g., Echidna/Medusa); formal verification (e.g., Certora/Satora). Use test harnesses to exercise edge cases.
  • System properties to internalize: atomicity of EVM transactions (partial failure reverts all), finality considerations, bridge/cross-chain assumptions.
• Continuous learning and adaptation:
  • Example: Tackled a complex system that combined contracts, protocol logic, and ZK components. Closed personal knowledge gaps in ZK by combining materials, targeted code reading, and AI-assisted exploration, yielding two distinct ZK vulnerabilities in a 2025 contest.
  • Takeaway: New tech surfaces constantly; adapt quickly, leverage AI and resources, and go where others aren’t looking.

Blockchain development: responsibilities, skills, and mindset (Hamid)

• Scope of work: Translate product/business logic into secure on-chain logic. Typical concerns include determining value flows, asset custody, and invariant enforcement (e.g., swaps, reserves, fee paths).
• Security mindset: Assume global adversarial scrutiny. Every line of code is public; bake in security reviews, pattern discipline (e.g., CEI), and defensive design from day one.
• Foundational prep for beginners:
  • Understand why blockchain exists and which problems it solves.
  • Learn smart contracts (semantics, lifecycle, tooling) and chain-specific nuances.
  • Build progressively and map logic to code and tests.
• Common beginner mistake: Learning in isolation. Better outcomes come from collaboration—pairing, code reviews, cohort-based learning, and building together. Community accelerates understanding and exposes you to professional practices.

Smart contracts vs. blockchain (clarification from Q&A)

• Smart contract development and blockchain development are distinct:
  • Smart contract development: Writing application logic deployed on existing blockchains (e.g., Uniswap contracts on Ethereum). It’s the “backend” of DApps.
  • Blockchain development: Building the underlying blockchain itself (e.g., Ethereum, Solana, L2s). Different algorithms, consensus, and infrastructure concerns. Smart contracts run on a blockchain; they are not the blockchain.

AI and Machine Learning

Anderson’s path and philosophy

• Background: Computer Science; strong math grounding; mentored by peers/senior colleagues. Progressed from Java/Kotlin to broader software and ML.
• Driver: Curiosity and disciplined inquiry (asking the right questions), more than money. Curiosity fuels sustained growth and deeper understanding.

AI vs. ML — simple distinction and professional reality

• Artificial Intelligence (AI): Broad field that creates systems mimicking human cognitive abilities (perception, problem solving, language, decision-making). Includes rule-based systems, robotics, NLP, and ML.
• Machine Learning (ML): A subset of AI enabling systems to learn from data without explicit programming—statistical modeling to recognize patterns, make predictions, and improve over time.
• Mental model: The “brain” is the ML model; AI is when you deploy that brain to autonomously act on real problems in products.
• Professional practice:
  • It’s not “movie magic.” It’s engineering: data collection/curation, model training/evaluation, deployment/serving, monitoring, iteration.
  • Strong software engineering is critical for ML engineers (integration, performance, observability, reliability).
• Examples mentioned:
  • Recommendation systems (e.g., e-commerce, Spotify cold-start lists).
  • Self-driving component stack (perception, decision, control).
  • Virtual assistants (Siri, Google Assistant).
  • An “AI writing detection” classifier project trained to flag LLM-generated text.
  • Personal exploration: “creating a brain” (learning-by-building mindset with NVIDIA tooling).

How to start and progress (Anderson)

• Pick your version of “hard” and commit. Every worthwhile path is challenging; consistency beats novelty-chasing.
• Avoid tutorial hell: Learn-by-building. When stuck, research a minimal fix, then return to the build.
• Don’t chase every shiny tool: Master fundamentals and one stack/language before branching.
• Build in public: Use GitHub; ship small projects; document clearly (docs win). Ask peers to review.
• Ask better questions: Do background research; show effort; be precise. People help when you ask well.
• Speed and iteration: Learn early, fail early, adapt quickly. Execution outpaces idea-hoarding.

Narrowing focus and team concerns (Q&A)

• For CS+Math students overwhelmed by breadth:
  • Start at an arbitrary point with a small, interesting project. Curiosity will guide the next steps.
  • Surround yourself with practitioners (mentors/peers) and ask for feedback; course correct as you go.
  • On “idea theft” anxiety: Ideas are abundant; execution and speed matter more. If it’s worth stealing, it’s worth building; vision and follow-through differentiate outcomes.

AI Engineer role and roadmap (Timilani)

• Why AI Engineering emerged: Post-2022/23 LLM boom (GPT, Gemini, Grok, etc.), demand grew for engineers who can build on top of foundation models and tailor them to business contexts.
• Practical AI Engineering example:
  • LLMs can hallucinate. An AI Engineer mitigates this by grounding the model with domain data (e.g., user-provided documents) via retrieval and conditioning, so outputs are accurate for the intended context.
• Suggested learning progression:
  1. Programming (commonly Python) and CS basics.
  2. Data Science (analysis, feature work, communicating insights for decisions).
  3. Machine Learning fundamentals (supervised/unsupervised, algorithms, evaluation).
  4. NLP (working with text, tokenization, embeddings, retrieval, generation).
  5. Deep Learning (neural nets as layered representations; practical frameworks; don’t be intimidated by the term “deep”—it refers to layered architectures).
  6. AI Engineering (RAG patterns, prompt design, evaluation, safety/guardrails, serving/integration, latency/cost/reliability trade-offs).

Standing out in AI (beyond technical skills)

• Ship projects and maintain a portfolio (GitHub preferred; if not, keep a well-organized Drive). Certificates are less persuasive than working demos and clear documentation.
• Learn fast by closing gaps quickly: if a topic is new in class, research it the same day, implement something small, and share it.
• Employers value demonstrable impact (how your project solves a real problem, speed-to-insight, and clarity of communication).

Community, program, and logistics (host notes)

• Community matters: Peer learning sustains motivation, exposes opportunities, and accelerates growth. Alumni groups and build events extend learning beyond classes.
• Applications: Deadline emphasized; selection emails expected next week; course-switch window will open in the community channels.
• Replays: Shared in community; scrub forward if silence at the start due to a 15-minute buffer before sessions begin.

Q&A highlights

• Cybersecurity career transition: Not a focus of this session (dedicated cybersecurity session referenced).
• Smart contracts and blockchain: Smart contract ≠ blockchain (see clarification above). Smart contracts are application logic on a chain; blockchain development is building the chain itself.
• Digital marketing path: Covered in a prior webinar; recording available via program channels.
• “Which AI areas will the boot camp cover?” Course outline will be shared with applicants next week.
• Model comparisons: A question about a specific model vs a “Claude”-family model; the speaker noted they did not have sufficient hands-on time before one model was taken down to provide a definitive performance comparison.

Closing reflections — what tech has done for the speakers

• Hamid: A sense of purpose and tangible impact—building useful systems and educating others.
• Timilani: Early opportunities (even during university), collaboration with professors (theory meets implementation), and strong referrals; reminder that different foundation models excel in different domains—choose tools by fit, not hype.
• Anderson: Fulfillment from doing work that matters to people; the importance of shipping things others can use.

Key takeaways

• Web3 is maturing: Less hype, more real applications (tokenization, privacy, ZK). Fundamentals + security mindset are non-negotiable.
• Web3 careers are diverse: Technical and non-technical roles exist; understanding blockchain basics enables skill transfer.
• Security research is deep work: Master fundamentals, study attack patterns, use the right tools, and keep learning—AI can accelerate your ramp-up.
• Smart contract vs blockchain dev: Different problem spaces and skill sets; don’t conflate them.
• AI vs ML: ML models are the “brains”; AI is the application of those brains to solve real problems. Software engineering skills are essential for production ML/AI work.
• Beginner strategy: Avoid tutorial loops; build publicly; focus on fundamentals; ask better questions; move fast and iterate.
• Roadmap to AI Engineering: Python → Data Science → ML → NLP → Deep Learning → AI Engineering (RAG, LLMOps, safety, serving).
• Portfolio over paper: Projects and documentation persuade more than certificates.
• Community compounds growth: Collaboration, feedback, and alumni ecosystems open doors and sustain progress.