Remove AI summaries and semantic searches

We were not using either. We can restore if we revisit.

app/models/concerns/searchable.rb

@@ -1,21 +1,6 @@
 module Searchable
   extend ActiveSupport::Concern
 
-  included do
-    has_one :search_embedding, as: :record, dependent: :destroy, class_name: "Search::Embedding"
-
-    after_create_commit :refresh_search_embedding_later
-    after_update_commit :refresh_search_embedding_later
-    after_destroy_commit :remove_search_embedding
-
-    scope :search_similar, ->(query) do
-      query_embedding = Rails.cache.fetch("embed-search:#{query}") { RubyLLM.embed(Ai::Tokenizer.truncate(query)) }
-      joins(:search_embedding)
-        .where("embedding MATCH ? AND k = ?", query_embedding.vectors.to_json, 20)
-        .order(:distance)
-    end
-  end
-
   class_methods do
     def searchable_by(*fields, using:)
       define_method :search_fields do
@@ -55,12 +40,6 @@ module Searchable
     update_in_search_index
   end
 
-  def refresh_search_embedding
-    embedding = RubyLLM.embed(Ai::Tokenizer.truncate(search_embedding_content))
-    search_embedding = self.search_embedding || build_search_embedding
-    search_embedding.update! embedding: embedding.vectors.to_json
-  end
-
   private
     def create_in_search_index
       fields_sql = [ "rowid", *search_fields ].join(", ")
@@ -95,12 +74,4 @@ module Searchable
       self.class.connection.execute self.class.sanitize_sql(statement)
       self.class.connection.raw_connection.changes.nonzero?
     end
-
-    def refresh_search_embedding_later
-      Search::RefreshEmbeddingJob.perform_later(self)
-    end
-
-    def remove_search_embedding
-      search_embedding&.destroy
-    end
 end