One Year Syllabus for Learning Software and A.I. Development "MSc"
This year, I will be mostly learning about Artificial Intelligence.
Spending a year and a half re-educating myself as a Cognitive Scientist / Ai Engineer.
20 years ago (2001-2004) (!!), I studied a Cognitive Science BSc. (Hons) at Exeter University in the UK.
Some interesting aspects of my BSc. Cognitive Science (Hons.)
The Blickets Paper (Developmental Science)
The first interesting thing about this course was that I ended up doing an internship at Carnegie Mellon University with Professor J. L. McClelland, who was a member of one of the founding groups of connectionist science in the cognitive sciences.
As a result of that summer in Pittsburgh, we produced an academic paper In the journal Developmental Science, which was quite an unusual thing to accomplish for an undergraduate student.
Here's the link to the paper if you're interested.
It was an interesting connectionist model illustrating a potential mechanism for how children reasoned about different causal events. The network used an interesting method of storing short-term memory in a simulated hippocampal set of units and then retraining the wider cortical network.It was essentially a highly simplified model of how the brain may process events from short-term memory into the long-term memory, which we know is a process that happens during sleep.
The Language Model
It's unfortunate that I wasn't able to preserve the original neural network model I constructed as my final year project. The network was largely hand-coded in C++, Using a lot of the knowledge that I gained over the preceding summer at the CNBC. It was, in essence, an extremely early language model based on a recurrent network that was able to very roughly simulate findings from eye movement studies based on the linguistic complexity that was being processed.
I trained a very simple network (by today's standards) to extract the grammatical structure from a simplified corpus of the English language. The argument was that delays in processing speed, as measured by eye movement statistics, were reflective of the statistical ambiguity based around the grammatical structures inherent in the English language.
I was impressed that this project was given 100% by my supervisor at the time at Exeter University, Who was a well-respected and reasonably well-known psycholinguist. When I visited his office after receiving my final grade, he said that he hadn't given me that grade because the project was perfect, but merely that I had done more than could reasonably have been expected by any other student in my position!
The End of Academia (for me, at least)
While I did have a talent for programming and thinking of academic things in an academic way, by the time I finished Uni, I was a bit burnt out and had had enough of sitting in front of a computer screen. Despite the fact that I achieved a First Class and Dean's Commendation, and was "heavily encouraged" by my lecturers to continue in academia and go on to do a PhD., I decided to "go my own way" (for better or worse). Since then**, I haven't done much programming, but I have learned a bunch of different stuff "the hard way" and had a wide variety of jobs and careers.
While most of these (Personal Training, Business Consulting, Carpentry & Joinery) involved a fairly short period of education, followed by mostly on-the-job training and learning, it seems a career in Artificial Intelligence and Software Development might take a bit longer up-front.
So I've developed a year-long syllabus for the self-taught "Master's Programme" I'll be putting myself through. It turns out I can keep myself fed and accomodated for about the same price I'd pay for an actual MSc, for roughly a year (in the end it took me 19 months to finish my studies (the culmination being the publishing of the Vox Manifestor App). So after selling almost all my possessions and moving to a tropical island somewhere in the Pacific Ocean, I'm ready to begin round 2 of my higher education journey.
Over the next year I will also be producing a variety of software in an online portfolio to showcase the skills and knowledge I am learning.
** notwithstanding the last year: I’m retrospectively posting this after 18 months of re-learning programming!!
MSc. in Artificial Intelligence and Software Development
(Self-designed / self-taught)
A comprehensive year-long program designed to equip the student for a career in the development of software and A.I. applications, with a robust understanding of the technical, conceptual, practical and philosophical implications in the development of advanced systems that are operated by machine intelligence.
By merging software development with AI specialization, we will develop the skills to innovate software solutions that integrate AI technologies, including autonomous systems, while also emphasizing data security and ethical practices.
The goal is to build the skillset and knowledge to produce software that harnesses and builds on the capabilities of AI, across multiple domains, in particular:
Human-computer-interaction,
Natural language and speech processing / production,
A.I. agents & assistants,
Desktop, Mobile and Web applications
Cybersecurity.
Module Guide
The program is structured into five key modules, each focused on critical aspects of AI and software development:
PURE_AI: AI theory, practice, and ethics.
MACHINE_LEARNING: Machine Learning foundations.
SOFTDEV/LANGS: Software development & programming languages / backend.
SOFTDEV/MOBILE: Software development & mobile applications / frontend.
CYBER_SECURITY: Privacy, identity security, cyber security, and data security.
Within these modules, you'll delve deep into various subjects, building a strong foundation that will enable you to excel in the AI and software development industry.
PURE_AI
Philosophy & Alignment
Learn about the philosophical implications and practical applications of AI alignment and AI safety.
Explore the roadmap for the harmonious integration of AI into societal structures.
AI Application Development
Master the principles of machine learning / neural network design and training, software design and the latest developments in AI technologies.
Stay updated with the latest in AI: ChatGPT, OpenAI's language models, image and speech recognition technologies, neural network architectures, and bolt-ons.
Design and implement cutting-edge interfaces for future human-computer/AI interaction, including the integration of autonomous agents.
MACHINE_LEARNING
Understand the key approaches and applications of AI, with a strong emphasis on cybersecurity and data protection.
Create custom language models and speech recognition models.
Develop machine learning models that prioritize security and privacy.
SOFTDEV/LANGS & SOFTDEV/MOBILE
Learn to write software for the backend, frontend, web, and mobile, with a focus on cybersecurity best practices.
NO CODE / LOW CODE Programming - how it works and what it is.
Develop and bring to market your own multi-user/community AI-based application, with an integrated desktop, web, and mobile implementation.
Creating detailed application design blueprints, adhering to security-by-design principles.
Understand and implement cloud computing solutions with a strong focus on security.
CYBER_SECURITY
Build a foundational understanding of cybersecurity, including the protection of personal data, and navigating risks associated with AI systems and autonomous agents.
Delve into human-computer interaction design, leveraging multi-constraint-based principles that incorporate gestural and voice recognition for secure interactions.
Develop applications that adhere to the highest standards of cybersecurity, from applied cryptography to secure coding practices.
Learn to proactively identify and remediate software vulnerabilities, including those in autonomous systems, through ethical hacking and penetration testing.
Formulate comprehensive cybersecurity strategies for AI-powered software solutions, with ongoing consideration for privacy and ethical implications.
Deliverables
Upon successful completion of the program, you will have achieved the following deliverables:
Open Source GitHub Portfolio:
Development of a portfolio of self-contained open-source tools, applications or libraries, in the areas of artificial intelligence, speech recognition and production, language models and language processing, autonomous agents, cybersecurity and / or privacy.
Portfolio projects now include detailed documentation that illustrates understanding of cybersecurity best practices, machine learning model development, and AI integration into full-stack development.
Demonstrations of GUI design proficiency and cloud computing deployments, with an emphasis on AWS, GCP, or Azure for AI-driven applications.
Application:
A fully operational mobile app with integrated AI features such as speech and image recognition, along with detailed security measures implemented (e.g., encryption, secure authentication).
Evidence of user testing, feedback incorporation, and iterative design for the mobile app.
Blog/Online Presence:
A series of in-depth articles on machine learning concepts, data science techniques, and their applications.
Regular updates on project progress, lessons learned from cybersecurity challenges, and AI ethics considerations.
A well-maintained blog that showcases thought leadership in AI concepts, systems, safety and alignment.
Social Media Presence/Strategy:
A tailored content strategy showcasing expertise in AI and machine learning, including participation in relevant online communities and discussions.
Building a following by sharing insights on cybersecurity trends, AI breakthroughs, and software development tutorials or tips.
Ongoing Paid Employment:
Leveraging the capstone project experience, including a comprehensive cybersecurity strategy and AI integration, in job applications and interviews.
Display of a strong understanding of cloud platforms, which are highly sought after in the industry, as demonstrated by cloud-related projects.
Additional Deliverables:
Machine Learning Model Documentation: Comprehensive documentation for each machine learning model developed, including data sources, algorithm choices, training processes, and version control.
UI/UX Portfolio: A collection of responsive web and mobile application designs showcasing a user-centered approach.
Ethical Hacking Certification: If applicable, a certification or badge from completing ethical hacking challenges, underscoring cybersecurity skills.
Kaggle Competition Participation: Evidence of participation in Kaggle competitions, along with any awards or recognitions.
These enhanced deliverables not only prove the student's capability but also enrich their professional profile, making them more attractive to potential employers or clients in the fields of AI, machine learning, and cybersecurity.
Topic & Module Breakdown
PURE_AI
Theory & Philosophy
AI Ethics and Societal Impact: Delving into the ethical and societal implications of AI technology.
Philosophical Implications and AI Safety: Discussing the philosophical underpinnings and safety considerations of AI.
Application
Large Language Models: Understanding, building, and applying transformers and LLMs.
Autonomous Agents: Development and application.
Robotics: Fundamentals and practical engagement.
MACHINE_LEARNING
Mathematics for neural networks and LLMs.
Machine learning concepts & algorithms / Data Science.
Speech recognition & voice production.
Custom Language Models: Techniques for creating and fine-tuning custom language models.
SOFTDEV/LANGS
Python programming.
GUI Design with Python.
Databases: PHP/SQL/NoSQL.
Web development, app integration: HTML, CSS, JAVASCRIPT.
Cloud Computing Foundations: Introduction to cloud platforms and services, emphasizing AWS, GCP, or Azure for AI development.
SOFTDEV/MOBILE
Design Blueprints and Mock-ups: Training on creating detailed application design blueprints and realistic mock-ups.
Android app development
Responsive Web Applications: Techniques for creating web applications that adapt to various devices and screen sizes.
User-Centered Design: Focusing on user interface design and user experience.
CYBER_SECURITY
Basic penetration testing & securing mobile devices.
Writing secure code and hacking applications
Protecting personal data
Applied cryptography, blockchain and the future of personal finance
Digital Footprints and Tracking: Understanding and managing online behavior to maintain privacy.
Midjourney 6.2: “a university student in the far future, studying advanced interfaces with droids and advanced computer programming interfaces all around --ar 40:21 --s 50 --v 6.1”
