TLDR;
This Instagram post by lostandlucky discusses Semantic Kernel and its applications, particularly in creating agents using structured files. The post highlights that many people are unfamiliar with Semantic Kernel, despite its potential. The comments section shows a range of questions and insights from users, discussing practical applications, the use of open-source models, and the impact of structured files on agent performance. The author promotes a free community for those interested in learning more about building in this space.
- Semantic Kernel is relatively unknown but powerful.
- Structured files can be used to define agents.
- A community is available for learning and collaboration.
Sjuul on Instagram
This section encourages viewers to follow Sjuul on Instagram to stay updated with their latest content. It also provides links to sign up for a new account or log in to an existing one.
Lostandlucky's Post on Semantic Kernel
Lostandlucky shares their experience working with Semantic Kernel since its release, noting that most people are unaware of it. The post includes a link to the original audio and options for further interaction.
Comments and Discussions
The comments section contains various questions and insights from users regarding Semantic Kernel and its applications:
- maracanimichele asks for a practical use case to understand the claims made about Semantic Kernel.
- stratusads inquires about using folder/file structures and Obsidian as a "brain" and how Claude Code interprets instructional markdown files.
- lucavalero questions whether an open-source model can be run locally using the described strategy without paying large language model (LLM) companies.
- jamesnminner seeks insights into how individuals are building and using their own agents in businesses and various applications.
- van_going expresses a desire to implement this at work but is unable to do so in their current role.
- healingwithjad asks if someone with no coding experience can use Semantic Kernel.
- anniesama8 describes Semantic Kernel as an optimised search engine that automates and processes data in a non-linear way, associating complex tasks and linking data with a specific direction towards "truth".
- brendan simply states "ontology. Palentir knew," possibly referencing the connection between Semantic Kernel and knowledge representation.
- mattr78 asks if there is an equivalent for Gemini.
- lookupjax affirms the post's content, sharing their Claude code workshop on GitHub, which taught them how to manage team leads, agent teams, and sub-agents from spec-driven markdown files.
- generoso_nyc questions whether recent research indicates that agent.md type files can negatively impact agent performance.
- loganbarker_ requests a video explaining the concepts in more detail for those new to development.
- rodrigowfranco shares a similar experience and concludes that simplifying the architecture is a good option.
- thesavedsalesman asks about the process of procuring data and storing knowledge into files.
- kingalethea enjoys learning from the discussions in the comments.
Additional Posts from Lostandlucky
This section showcases more posts from lostandlucky, each promoting their community and sharing insights on AI, business, and software development:
- A reminder that experience isn't everything and encourages focusing on the bigger picture of AI.
- An invitation to join a free community for learning and collaboration.
- A discussion about financial outcomes and the efficiency of their systems.
- An answer to a question, likely from a previous post or comment.
- Advice to focus on what is valuable to people rather than automating everything.
- An invitation to join a community of over 12,000 people learning about software and business fundamentals.
- An offer of support and resources within the community.
- An announcement about sharing repositories, code bases, and zip files.
- An invitation to join a community to learn more about the fundamentals.
