It seems that there will be many products and forms of agent in 2025. Why does it seem...

It seems that there will be many products and forms of agent in

Why does it seem that only coding agent has emerged? One perspective is: Is AI Agent a technical term or a product form? Even at the end of 2025, the current situation is still technically called this, and the product naming is also called this. Even the word Agent itself has been around for a long time. If you look at the AI Agent in 2025 from the technical side, it has actually made great progress:
The AI Agent has improved the ability and accuracy of tool integration, and the basic model side has basically been specially trained with agentic capabilities to adapt to the AI Agent.
Contextual engineering has gone from concept to engineering implementation to large-scale popularization in a short time, which has greatly accelerated the control ability, accuracy, and memory ability of the AI Agent when interacting with large models.
Agent Infra has also made great progress, and now the Agent infrastructure can support hundreds of sub-agents. Activation and release in a short time 4 The Coding Agent capabilities mentioned in this post can be divided into model-side training that can be divided into front-end UI related code capability improvement (images, interfaces, PPT...), voice TTS, back-end code pre-training, and improvement of mathematical capabilities (in fact, mathematical capabilities are very important for coding) 5 AI Agent startups have made great innovations and progress at the application layer engineering level, so that the basic model in turn learns from the application side. A typical example is Claude learning manus @ManusA I launched Claude Skill, and now all major model manufacturers are launching their own Skills one after another. Therefore, it should be called the first year of AI Agent technology, which is more accurate. What I see through my eyes is huge technological progress.

Visual summary

Article argument map

Generated from the post's content graph

Mermaid outline

flowchart LR
  format-long_post["longform note"]
  topic-agents["agents"]
  topic-ai-workbenches["ai workbenches"]
  topic-memory["memory"]
  topic-tooling["tooling"]
  capability-agent-workflow["agent workflow"]
  capability-ai-native-workbench["AI-native workbench"]
  capability-product-surface["product surface"]
  capability-technical-distribution["technical distribution"]
  market-ai-startup["AI startup"]
  format-long_post -->|covers| topic-agents
  format-long_post -->|covers| topic-ai-workbenches
  format-long_post -->|covers| topic-memory
  format-long_post -->|covers| topic-tooling
  format-long_post -->|signals| capability-agent-workflow
  format-long_post -->|signals| capability-ai-native-workbench
  format-long_post -->|signals| capability-product-surface
  format-long_post -->|signals| capability-technical-distribution

Visual structure

Essay structure map

Built from summary and key paragraph positions

Mermaid outline

flowchart LR
  thesis["It seems that there will be many products and forms of agent in 2025. Why does it seem that only coding age..."]
  signal["It seems that there will be many products and forms of agent in"]
  operator["1. The AI Agent has improved the ability and accuracy of tool integration, and the basic model side has b..."]
  implication["3. Agent Infra has also made great progress, and now the Agent infrastructure can support hundreds of sub-a..."]
  thesis -->|frames| signal
  signal -->|develops| operator
  operator -->|lands in| implication

Source: View the original post