AI Integration Strategy for Learning and Knowledge Management Solutions
Acomparative study of 100 generative AI-based tools in the context of learning & knowledge management.
Artificial Intelligence (AI) technology and particularly Generative AI has gained mainstream attention, the public media, as well as the business corporation. The global AI market is valued at $ 142.3 bn as of 2023 and it’s expected to continue to grow into a $ trillion market in the coming years (Source). On another note, the AIAAIC has reported that the number of incidents and controversies concerning the misuse of AI is exponentially increasing from 123 incidents (2019) to 1,000+ incidents (first half of 2023). As Knowledge Management (KM) leaders and practitioners, it’s critical to have an active role in guiding the integration of Generative AI into KM areas, applications, and processes. This report is an attempt to provide some guidance on the current state of Generative AI integration within the KM context. Specifically, the report is answering the following question:
“Where and how Generative AI is accelerating and impacting knowledge use cases, areas, and processes?”
I have conducted a comparative study of 100 Generative AI tools in the context of learning and knowledge management that has resulted in a set of 35 KM processes where Generative AI has augmented their experience, implementation, and execution. You may access the evaluation grid here.
AI-Based KM Solutions by Service Area
AI-based KM solutions have served multiple areas ranging from market intelligence, sales prediction, customer service, and intelligent search to employee learning and development. The below chart shows the distribution of the 100 AI-based KM solutions by service area:
· Customer Services: customer experience, customer call center, customer feedback analysis, incidents, ticket management, self-service customer portals, and customer web chat.
· Cognitive Services: intelligent search, semantic search, cognitive search, symbolic search, and insights engine.
· AI Infrastructure for KM: AI and machine learning platforms, symbolic and reasoning algorithms, deep learning, large language models, and neural networks.
· Content and collaboration platform: intelligent document and process automation, content automation, content optimization, content editorial, content summarization, content intelligence, enterprise knowledge platform, knowledge sharing platform, and knowledge base.
· Sales and Marketing: Brand Experience, digital marketing, digital advertising, market intelligence, and research intelligence.
· Learning and Development: Smart Skills, Digital Worker, Talent engagement, and social learning platform.
AI-Based KM Features by Process Area
The study found 35 KM processes that have been promoted with Generative AI technology and features. We have organized the KM processes into 7 Knowledge activities:
- Social Learning & Personal Capabilities.
- Knowledge Co-Development & Exchange.
- Knowledge Retention & Reuse.
- Expertise Discovery & Dissemination.
- Knowledge Discovery & Generation.
- Knowledge-Centered Services.
- Knowledge Analytics & Intelligence.
Generative AI features were directly impacting 35 KM processes and they are as follows:
1. Social Learning & Personal Capabilities
1.1. Personalized Learning & Assisted Coaching
1.2. Skills Suggestions for Capabilities Building
1.3. Automated Multi-Language Learning
1.4. Dynamic Creation of Curated Learning Style
1.5. Suggestion of Matching Mentor-Mentee Pairs
2. Knowledge Co-Development & Exchange
2.1. Suggested Topical Communities: Problem-Solving
2.2. Augment Knowledge Sharing Behaviors
2.3. Digital Project-based Workspace
2.4. Knowledge Narratives and Storytelling
2.5. Ideation and Collaborative creativity process
3. Knowledge Retention & Reuse
3.1. Assist in Creating Rich Content (Wiki, KB, Reports)
3.2. Feedback Loops and Lessons Learned
3.3. Content Syndication & Inline Integration
3.4. Insights and Best/ Direct Answers Extraction
3.5. Knowledge Portal: Diffusion of reusable content
4. Expertise Discovery & Dissemination
4.1. Human-In-The-Loop Collaboration (Experts)
4.2. Assist in creating people’s profiles.
4.3. Inference of Expertise and micro-skills
4.4. AI Digital Worker & Pre-Built Smart Skills
4.5. Sentiment/Intent-aware recommendations
5. Knowledge Discovery & Generation
5.1. Content Views / Topic-based Dashboard
5.2. Content Optimization (Accuracy, Clarity…)
5.3. Content Automation & Operational efficiency
5.4. Recommend Relevant/ Confident Content
5.5. Content Re-Generation & Conversational Search
6. Knowledge-Centered Services
6.1. Customer Conversations & Ticket Resolution
6.2. Brand Community & Customer Engagement
6.3. Self-service Portal and FAQs Suggestions
6.4. Automatic Routing & Real-time Agent Assistance
6.5. Customization and API Integration with systems
7. Knowledge Analytics & Intelligence
7.1. Meta-Data Enrichment (Classifier, Dictionary…)
7.2. Generation of Social/ Knowledge Graph
7.3. Content analysis capabilities (predictive, NLP)
7.4. Content Reporting & Real-time Data Visualization
7.5. User, Community & Business Analytics
Summary & Conclusion
The Upside
Generative AI can help in personalizing the learning experience, provide assisted coaching, and suggest skills for individual career growth. AI can help in curating and suggesting relevant learning materials in multiple languages. In addition to transcribing and translating audio & video content for almost instant access for the global audience. AI can also help in drafting missing articles, reports, and creating knowledge portals. They can also augment the community experience by suggesting topical communities, matching mentor-mentee pairs, enriching the sharing behaviors with sister communities, and generating knowledge narratives, and stories. AI can help in accelerating the ideation and creativity processes by mapping and connecting ideas and idea authors to the innovation campaign objectives. AI can assist in creating rich content either by drafting missing knowledge base articles or by auto-completing ideas with arguments and examples. AI can search and extract answers from specific document sections after analyzing the context, the sentiment, and the intent of the user query. AI can help in improving human-the-loop collaboration by suggesting experts based on their activities, involvement, and preferences. AI can also infer expertise and micro-skills by analyzing content authors and contributors’ behaviors and patterns of engagement. As consequence AI can assist in augmenting peoples’ profiles with micro-skills and topics of interests. AI has introduced the concept of digital worker who’s a human like clone for specific roles, responsibilities, and tasks. Digital workers are characterized by being curious, collaborative, and capable. Digital workers can be pre-loaded with smart skills which are signature patterns and workflows replicating industry-specific practices. AI can automate, extract and regenerate knowledge into new formats or schemas. They can extract the competitor’s website and all the spider’s links into a structured data list with their properties and meta-data. AI can also enrich content with meta-data and semantic relationships and generate a text into a knowledge graph heavily relying on natural language processing (NLP). AI can impact the customer’s experience by offering self-service portals, FAQs or recommend relevant content making the customer experience more conversational and interactive. AI can integrate with multi-modal systems to scale the data infrastructure for a more comprehensive and integrated database by linking it to different sources of information. AI can automate the data migration process by removing duplicate entries or by combining similar data.
The Downside
AI algorithms require guidance and supervision to define which information and data are most important to the users to lay down the foundation for meaningful insights and best/ direct answers. It’s important to note that while large language models like GPT-3 are powerful in generating text, they are not inherently intelligent or conscious. They don’t possess an original understanding or awareness of the content they generate and may sometimes produce outputs that are nonsensical or inappropriate. Therefore, they require human supervision — known as reinforced learning and careful application to ensure their outputs meet the desired outcome. The cost of implementing an internal infrastructure for Generative AI/ LLM technology deployment can rise significantly based on factors such as the scale of deployment, the complexity of the technology, the size of the organization, and the specific requirements of the project. On the contrary, if an organization decides to use cloud based LLM cognitive services, concerns around governance, security and privacy will be a legitimate subject to carefully consider.
