The Semantic Web project didn’t fail. It delivered—early. What it placed on the table remains unmatched as foundational infrastructure for solving critical problems. What was missing wasn’t vision, but timing. LLMs have now arrived. Which means a Semantic Web-driven evolution of the Web, as […]

I do a lot of hands-on work with AI Agent Skills that are loosely coupled with Data Spaces (databases, knowledge bases, filesystems, and APIs). OpenLink Software now has a publicly accessible AI Agent Skills repository on GitHub that I contribute to directly […] [Original post on mastodon.social]

RE: https://mastodon.social/@kidehen/116302922748631627 The links in that important announcement have been fixed i.e., they no longer return 403's. @egonw: Thanks for the heads-up!

Love how https://consensus.rubalkhali.science/ links ecological data to the molecular world of #uniprot with some smarts and the power of federated #sparql

At a pivotal moment—when AI agents are rediscovering and reinforcing the enduring value of Linked Open Data (LOD) and the principles of Linked Data—we are taking another step forward for the community. We will soon release a new open-source edition of Virtuoso that introduces a feature […]

Here’s a Feed Generator Skill I knocked up that generates feeds for pages that lack this important content syndication and discovery feature. How do you use it? From any AI or AI agent environment that supports skills packaged using SKILLS.md docs. Read on..

LLMs + the Model Context Protocol (MCP) are the Yang to the Semantic Web Project's Yin. We now have a solution to the final hurdle—visualization. Years of Linked Data work now come alive. I explain this, with demonstrations, in a new newsletter post. www.linkedin.com/pulse/semant... #mcp

The Model Context Protocol (MCP) brings Unix-style loose coupling to AI Agents. I built a visual explainer (JSON-LD metadata included) using Google Gemini’s new Canvas feature—with an embedded Knowledge Graph to illustrate. Check it out […] [Original post on mastodon.social]

A simple GUI-based screencast demo of the @openlink AI Layer (OPAL), showing what’s now possible via MCP + A2A—namely, composing and orchestrating agentic workflows through loosely coupled software. How? An Agent Host routes tasks to Agents via A2A. Agents […] [Original post on mastodon.social]

An example of what’s now possible—thanks to our (@openlink) sophisticated MCP Server for JDBC, Oracle’s native JDBC Driver, and Anthropic's Claude Desktop as the MCP client: A dashboard generated from Oracle’s HR database using the following natural language […] [Original post on mastodon.social]

This week’s edition of my @linkedin newsletter shines a light on the impact of the Model Context Protocol (MCP) across multiple computing eras—leading right up to today. We’ve arrived at a moment where loose coupling principles are cool (and super useful) […] [Original post on mastodon.social]

The Model Context Protocol acts as a universal translator between LLMs and data sources, eliminating complex platform-specific requirements. Our new open source MCP server for ODBC (mcp-odbc-server) enables seamless integration of any ODBC-accessible […] [Video] [Original post on mastodon.social]

@medium I've requested a data export over 24 hrs ago. I am still awaiting email notification for the download link.

How can we prepare for a future migration from Bluesky to Mastodon? Bridgy Fed — a service that enables you to connect together your website, fediverse account and Bluesky account — will help. But Bridgy Fed needs to be easier to use. So I recruited Claude’s new Sonnet 7 model to help me do […]

LLMs are the “Easy Button” for understanding and harnessing the powerful benefits envisioned by The Semantic Web Project. If you closely examine the GenAI disruption in software, you’ll notice that all roads lead to better note-taking—progressively enriching knowledge repositories that […]

Here’s a DataTwingler AI Agent configuration in Markdown, designed for interacting with UniProt’s massive Knowledge Graph, which consists of 210 billion machine-computable statements of fact, meticulously curated by domain experts. github.com/OpenLinkSoft...

Here’s a DataTwingler AI Agent configuration in Markdown, designed for interacting with UniProt’s massive Knowledge Graph, which consists of 210 billion machine-computable statements of fact, meticulously curated by domain experts […] [Original post on mastodon.social]

Links: * https://sparql.uniprot.org/.well-known/void -- Knowledge Graph Stats * https://sparql.uniprot.org/sparql -- Query Service Endpoint * https://github.com/sib-swiss/sparql-training/tree/master/uniprot -- SPARQL Querying Tutorial * […] [Original post on mastodon.social]

UniProt’s latest publicly accessible Knowledge Graph now contains 210,122,358,019 RDF triples—powered by our (@openlink) #virtuosordbms open-source edition, live on the Web and available to AI Agents, Applications and Services across life sciences, genomics, bioinformatics, molecular biology […]

UniProt’s latest publicly accessible Knowledge Graph now contains 210,122,358,019 RDF triples—powered by our (@openlink) #virtuosordbms open-source edition, live on the Web and available to AI Agents, Applications and Services across life sciences, genomics, bioinformatics, molecular biology […]

The pendulum is swinging back toward the edge of the global connectivity mesh we call the #Internet—a familiar cycle, much like the shift from time-shared mainframes to the PC and Mac era. Microsoft's Phi-4 LLM running on my MacBook Pro (M1), courtesy of LM […] [Original post on mastodon.social]

Quick evaluation test screencast and a tabulated offer analysis for the new GPT-o3 LLM family from #openai. #genai #ai

The latest edition of my newsletter is out! This week, I focus on a deceptively simple reasoning and inference challenge to compare the capabilities of AI+Compute inference services from #OpenAI, #Google, #Microsoft, #Groq, #Alibaba (Qwen), #DeepSeek, and […] [Original post on mastodon.social]

NotebookLM was impressive, but now we have the visual edition from HeyGen.ai Check out this fantastic video podcast on the OpenLink AI Layer (OPAL)—an add-on to our Virtuoso Data Spaces platform. https://youtu.be/ceBgsOW2EQI #ai #genai #howto #knowledgegraph #opal #video #podcast

Q: Do large language models (LLMs) possess reasoning and inference capabilities? A: Yes-ish. The issue is that “yes-ish” isn’t enough for scaling an AI + compute business model. LLMs will struggle to evolve toward AGI without integrating Knowledge Graphs via RAG or GraphRAG, where reasoning […]

Happy New Year! Check out the latest article in my newsletter: 2025 Technology Predictions: A Shift Towards Agentic Workflows and Decentralized Innovation. https://www.linkedin.com/pulse/2025-technology-predictions-shift-towards-agentic-workflows-idehen-inbne #trends #ai #genai

#Google Gemini 2.0 now passes my “deceptively simple” test: generating #SPARQL queries against the #DBpedia #KnowledgeGraph. This marks the first time Gemini LLMs have succeeded in my testing. Ironically, it’s less about core #LLM capability and more about […] [Original post on mastodon.social]

Boom! The OpenLink Software #AI Layer (#OPAL) now supports AnthropicAI's #Claude language models. See comments for a #DataTwingler AI Agent demo executing #SPARQL and #SQL queries via a live #VirtuosoRDBMS instance. Remember, query scope is SPARQL, #ODBC […] [Original post on mastodon.social]

@Grok (v2) just passed my #sparql query generation test on the #dbpedia #knowledgegraph! It can now generate a link resolving to all movies directed by Spike Lee from my test prompt. https://x.com/i/grok/share/mnGYuedRFavNWaQk4XxjUPVm1 Links are shareable too—this is HUGE for #genai & […]

I’m loving my timeline here in the fediverse. I seamlessly follow and engage with people posting via Mastodon, Threads, Bluesky, Pixelfed, Lemmy, Ghost, Wordpress and Flipboard. The social web is here and it’s just awesome.

And what’s also fun is I’m getting likes and replies from people using different apps and networks. So cool! Feels like Mastodon is the Rosetta Stone of the social web.

This is another #fediverse test. Hoping that I can route this to Threads, courtesy of the cc. /cc @[email protected], @[email protected]

#threads now supports #fediverse following. URL Pattern: https://threads.net/fediverse_profile/{fediverse-handle}. #mastodon Follow Example: https://threads.net/fediverse_profile/@[email protected] #howto #openweb #nosilo

Despite the rapid pace of technological innovation, one persistent challenge remains: the growing number of apps paired with a diminishing ability to comprehend their underlying data landscapes. This issue always challenges enterprise agility […]

This week’s newsletter: “ChatGPT Search == Semantic Web Easy Button!” 🔗 https://www.linkedin.com/pulse/chatgpt-search-semantic-web-easy-button-kingsley-uyi-idehen-a5doe/ The notion of a #semanticweb is one of the most misunderstood aspects of the Web’s evolution. Thanks to #llm innovations […]

This is the nightmare scenario where #ai has the potential to completely upend society as we know it! Consider this advanced notice—once again! https://joanwestenberg.medium.com/the-death-of-critical-thinking-will-kill-us-long-before-ai-781fdd23cc7c #society #socialmedia

New Qwen-QwQ-32 LLM passes my tricky #dbpedia #sparql query results link test. Prompt: Using the DBpedia endpoint, list all movies directed by Spike Lee and encode the sparql query in a clickable link. Result: dbpedia.org/sparql?query... Chat: https://huggingface.co/spaces/Qwen/QwQ-32B-preview

How to generate a powerful Knowledge Graph from a LinkedIn article using a Large Language Model (LLM) such as OpenAI’s GPT-4o or GPT-4o-mini. #ai #genai #usecase #llms #knowledgegraph #howto

Hyperlinks even bring powerful innovation to the realm of #DBMS technology, which has been devoid of real innovation for eons! I explain how in the following post: www.linkedin.com/pulse/addres...

What do I mean by FORCE unleashed by the #Web? I explained it years ago in a blog post. www.openlinksw.com/dataspace/do... #OpenWeb #Society

Hyperlinks are the key to harnessing the FORCE unleashed by the #Web. Any platform that obscures this power does so for short-sighted reasons, aiming to limit user freedom. Great platforms thrive by delivering quality service—not by resorting to enshittification-based lock-in. #OpenWeb #WebWeWant

The original sin of #Web20 was giving siloists a pathway (albeit short-sighted and ill-conceived) to colonize the #OpenWeb. This is evident even today in #AI chatbot interfaces that struggle to include shareable links or handle links effectively in their responses.

"Links represent a threat to closed systems." -- @anildash.com. Read: www.anildash.com/2019/12/10/l... #OpenWeb #SocialMedia

The latest edition of my weekly #newsletter is out! This week’s focus is on tackling ‘information overload’ in the age of #ai. https://www.linkedin.com/pulse/knowledge-graphs-meet-genai-turning-information-overload-idehen-lrm8e/ #genai #usecase #knowledgegraph #DataSpaces

This is what user-controlled identity, preferences, and profile information look like—entirely built on existing open standards to address the longstanding challenge of online identity authenticity. #youid #netid #webid #linkinbio #pds #DataSpaces

This week’s #newsletter is out. It explores the powerful synergy between #LLM-based #ai Agents and Knowledge Graphs. https://www.linkedin.com/pulse/ai-agents-knowledge-graphs-kingsley-uyi-idehen-nmr1f #genai #knowledgegraph #usecase #cdo #cio #caio #DataSpaces

New edition of my weekly #newsletter is out! This week’s focus is Verifiable #Identity & Personal Profile Pages using: 1. #Hyperlink-based Entity Naming 2. Profile-oriented Knowledge Graphs 3. Loose coupling with #AI Agents […] [Original post on mastodon.social]

The latest edition of my weekly #newsletter is out. This week's edition is about a customer support #AI Agent created declaratively using natural language. https://www.linkedin.com/pulse/creating-product-support-ai-agent-using-natural-language-idehen-ivcof […] [Original post on mastodon.social]

Another Test post via my #odsbriefcase magic folder.