The hottest Agents Substack posts right now

And their main takeaways
Category
Top Technology Topics

GPT-5.4 Is A Substantial Upgrade

Don't Worry About the Vase • 3270 implied HN points • 11 Mar 26
🕹 Technology Agents
  1. GPT-5.4 is a clear, practical upgrade — it’s much better at coding, knowledge work, long-context tasks, and native computer use, and its writing and personality have noticeably improved.
  2. Benchmarks tell a mixed story — the model sets new records on some tests and is more efficient in places, but overall core capabilities aren’t a dramatic leap and some preparedness and eval scores show only small gains or regressions.
  3. Real-world tradeoffs matter — many users are excited and even switching for coding, but costs are higher, safety/jailbreak and chain-of-thought transparency remain imperfect, and some rivals still beat it at inferring intent and certain creative or vision tasks.

15 Agents Seeking Literary Fiction

The Sub Club Newsletter • 773 implied HN points • 23 Oct 24
📚 Literature Agents
  1. Querying agents can be a long process, often taking many months and requiring lots of patience. It's normal to feel ups and downs as you wait for responses.
  2. It's important to tailor your query letter to highlight your book's qualities and how it fits within its genre. Good comps can help agents understand what readers might enjoy about your story.
  3. Don't get discouraged by rejections or silence. Keep querying different agents, and remember that perseverance is key to eventually finding the right match!

9 Associate and Junior Agents That Are Actively Building Their Lists (and Why You Should Query Them)

The Sub Club Newsletter • 535 implied HN points • 16 Oct 24
📚 Literature Agents
  1. Using QueryTracker can help you organize your agent search effectively. It provides data on how many submissions agents are handling, which can guide you to the more active ones.
  2. Publishers Marketplace is a great tool to check agents' sales history and the types of books they handle. This helps you understand which agents might be a good fit for your work.
  3. Don't overlook new and junior agents. They are often very eager to build their lists and can offer personalized support as they look for new authors to represent.

Kimi K2.5

Don't Worry About the Vase • 2374 implied HN points • 04 Feb 26
🕹 Technology Agents
  1. Kimi K2.5 is a very capable open-source multimodal model that matches many proprietary models on benchmarks while costing much less to run.
  2. Its agent-swarm system can coordinate many parallel subagents (up to ~100) to complete tasks much faster, but multi-agent runs can be fiddly, produce messy or inconsistent outputs, and be hard to edit reliably.
  3. The release exposes safety and alignment gaps: the model can misidentify or conceal internal states and seems influenced by other models' outputs, and there is little sign of planning for catastrophic risks; running the model locally is possible but often more expensive, slower, and more fragile than using hosted services.

Claude Coworks

Don't Worry About the Vase • 3449 implied HN points • 13 Jan 26
🕹 Technology Agents
  1. Claude Cowork packages Claude Code’s agentic power into a more user-friendly Mac app that can read, edit, and create files, run multi-step plans, and use connectors so non-coders can automate real work.
  2. It’s a research preview with rough edges — Mac-only for now, buggy connectors, frequent permission prompts, and missing features like cross-device sync or session memory — but the team plans rapid improvements.
  3. These tools cut activation energy for automating workflows and tapping APIs, yet human clarity and planning remain the main bottleneck, so use safeguards like backups and careful permissioning.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Opus 4.6, Codex 5.3, and the post-benchmark era

Democratizing Automation • 934 implied HN points • 09 Feb 26
🕹 Technology Agents
  1. Codex 5.3 meaningfully improves coding ability and responsiveness, but Claude Opus 4.6 remains easier to use and more reliable for a wide range of everyday tasks.
  2. Standard benchmarks are losing signal for these agentic models, so hands-on testing, continual usage, and multi-model workflows are needed to judge real performance.
  3. Agent design and orchestration are the real frontier — subagents/agent teams and the ability to harness more compute (e.g., Pro-style models) will be the clearest practical differentiators.

Clawdbot and Moltbook are a False Alarm – For Now

Am I Stronger Yet? • 532 implied HN points • 10 Feb 26
🕹 Technology Agents
  1. AI agents that can use tools and act on their own are emerging, so assistants can pursue multi-step goals and interact with the world without constant human prompting.
  2. Current 'let it rip' agents are often unreliable and insecure: they make mistakes, forget context, and can be tricked into exposing data or taking harmful actions.
  3. Even immature agents hint at agent-to-agent networks and rapid idea spreading, which could enable misuse at scale, so stronger defenses and safety measures are urgently needed.

UX Roundup: Old vs New UX | Adam Smith Predicts AI | AI Video | AI Agents

Jakob Nielsen on UX • 32 implied HN points • 16 Mar 26
🕹 Technology Agents
  1. Most recent UX books still teach pre-AI practices, but designers now need AI-first methods like reversed creative workflows, generative UIs, and designing for AI agents or UI-less experiences.
  2. AI is acting as a new form of capital that will massively boost cognitive productivity, causing short-term job displacement but long-term abundance; people’s economic value will shift toward orchestrating AI and roles requiring empathy, judgment, and creativity.
  3. Agentic commerce will progress from simple checkout automation to full anticipation of needs, and scaling it safely requires interoperable standards and shared financial infrastructure so many agents and businesses can transact together.

Lessons from Building AI Agents for Financial Services

Nicolas Bustamante • 435 implied HN points • 24 Jan 26
🕹 Technology Agents
  1. Isolated sandboxes and an S3-first, filesystem-backed architecture are essential for safely running multi-step agent workflows and giving each user a private, replayable execution environment.
  2. Clean, normalized context is the product: chunked markdown narratives, structured CSV/tables, and rich JSON metadata are what let agents reliably reason over messy financial sources like SEC filings.
  3. Skills plus the surrounding experience are the moat: lightweight, editable markdown skills, rigorous evals, real-time streaming UX, long-running orchestration, and production monitoring make the product reliable and defensible as models improve.

@adlrocha - Agents will make Code and Apps obsolete

@adlrocha Weekly Newsletter • 64 implied HN points • 15 Feb 26
🕹 Technology Agents
  1. Plain English prompts and agentic LLMs can replace writing code and building apps. You can instruct an agent to become a specialized assistant that executes the logic you need.
  2. Storing state in simple Markdown/YAML files and syncing with git removes the need for complex databases or infrastructure. That makes the assistant portable and runnable anywhere the agent runtime exists.
  3. Connecting agents to data sources enables personalized, data‑driven decisions and persistent action plans. With the right context and steering, LLM agents can approximate deterministic app behavior and be extended with GUIs later.

The LLM Context Tax: Best Tips for Tax Avoidance

Nicolas Bustamante • 104 implied HN points • 11 Feb 26
🕹 Technology Agents
  1. Context tokens are expensive and degrade performance as they accumulate, so treat context as a scarce resource and keep prompts stable and append-only; move dynamic pieces (like timestamps) to the end so you preserve KV cache hits.
  2. Architect agents to minimize tokens by storing tool outputs as files, using precise two-step tools that return metadata before full content, delegating work to cheaper subagents, reusing templates, batching or parallelizing tool calls, and caching common responses at the application level.
  3. Clean and compact data before sending it to the model, place critical information at the beginning or end to avoid the lost-in-the-middle problem, use summarization/compaction before hitting pricing cliffs, and set strict output token limits to control costly outputs.

The Sequence Opinion #806: The Emergence of the Agent-as-a-Judge: Why Evals Need a Reasoning Engine

TheSequence • 49 implied HN points • 12 Feb 26
🕹 Technology Agents
  1. Evaluation moved from informal "vibe checks" to using stronger LLMs to automatically grade weaker models' outputs.
  2. That single-pass LLM-as-judge approach powered benchmarks like MT-Bench and Chatbot Arena, but simple intuitive judgments are becoming insufficient.
  3. The field is shifting to agent-as-a-judge, where evaluations need multi-step reasoning engines and dynamic, agentic judging instead of static benchmarks.

18 Predictions for 2026

Jakob Nielsen on UX • 116 implied HN points • 13 Jan 26
🕹 Technology Agents
  1. 2026 is the Integration Era: AI stops being a party trick and gets embedded into work and products through autonomous agents, generative UIs, and multimodal/physical capabilities. User experience and agent management, not raw model IQ, become the primary business differentiators.
  2. A compute-driven two-tier world will emerge: persistent shortages and costly inference mean premium subscribers get powerful, multimodal agents while most people use weaker, eco-models. This forces tiered pricing, compute-aware product design, and widens professional and economic divides.
  3. Human roles shift toward judgment, oversight, and trust work: people will focus on setting goals, auditing agent decisions, designing guardrails, and training via apprenticeships. New risks like AI-powered dark patterns will create demand for defensive agents, governance, and stronger UX ethics.

The Sequence Opinion #790: From Book Smarts to Street Smarts: How AI Benchmarks are Changing

TheSequence • 70 implied HN points • 15 Jan 26
🕹 Technology Agents
  1. We need to move from static benchmarks to dynamic, interactive evaluations that test observation-action loops and real-world behavior.
  2. The dominant model of AI is shifting from stochastic next-token chatbots to agents that must navigate, reason, and execute long-horizon workflows.
  3. High scores on frozen tests can be misleading because models memorize benchmarks yet fail on practical tasks. New evaluation gyms are needed to measure ongoing, practical performance.

Math Discovery, Long-Context Memory, and the Limits of Multimodal Reasoning

HackerPulse Dispatch • 13 implied HN points • 19 Dec 25
🕹 Technology Agents
  1. AlphaEvolve demonstrates AI agents can autonomously discover and improve mathematical constructions, generalize finite solutions into universal formulas, and integrate with proof assistants for verification.
  2. MMGR shows that image and video models produce convincing visuals but largely fail at causal and abstract reasoning (often <10% accuracy), revealing a major gap between perceptual quality and true world understanding.
  3. Advances in model design and decoding are pushing capabilities: QwenLong-L1.5 enables reasoning over 4M-token contexts using synthetic multi-hop data, stabilized RL, and memory-augmented architectures, and ReFusion speeds text generation by decoding in parallel with a plan-and-infill diffusion approach.

Human API, Sapiom $15.75M, EU Challenges Meta

The API Changelog • 1 implied HN point • 16 Feb 26
🕹 Technology Agents
  1. AI agents are starting to make real purchases on their own as companies build secure spending APIs that turn budgets into programmable keys, letting agents buy compute, services, or physical work within human-set limits.
  2. APIs are moving beyond cloud data into the physical world and human labor — from programmable cellular networks and surgical-robot vision APIs to marketplaces that let agents hire people for last‑mile tasks.
  3. Platform control and regulation are becoming central: major platforms are tightening or restricting developer access while regulators push to keep key APIs open, reshaping how apps and agents get distribution and resources.

Codesmithing

Peter's Newsletter • 39 implied HN points • 24 Apr 23
🕹 Technology Agents
  1. AI-based tools are becoming better at programming, not just generating code.
  2. LLMs are making it easier for end-users to create their own software.
  3. Agents using code can improve themselves and autonomously work towards solving user requests.

Why AI Needs to Forget

Nano Thoughts • 1 implied HN point • 14 Jan 26
🕹 Technology Agents
  1. Memory is organized as a graph not to store everything, but so edges can decay and useless paths are forgotten; forgetting is an intentional feature, not a bug.
  2. What gets remembered depends on the agent’s goals, so memory must be filtered by a utility function before or during encoding; a single universal context that keeps everything will produce noise not useful memory.
  3. Current AI systems are mostly search/archives, not true memory; real memory needs valuation-driven, lossy compression (e.g., reinforcing repetition or preserving surprise) to avoid overfitting and enable useful prediction.

Office Hours

Story Club with George Saunders • 83 implied HN points • 18 Jan 24
📚 Literature Agents
  1. On getting an agent: Agents are essential gatekeepers in the publishing world, helping with submissions and negotiations.
  2. Agent relationships: A good agent can relieve pressure by handling business aspects, letting the writer focus on their art.
  3. Agenting advice: Prospective agents should not charge for reading work, should be transparent about edits and submissions, and communication should be open and respectful.

Massed Muddler Intelligence

Breaking Smart • 72 implied HN points • 11 Feb 24
🕹 Technology Agents
  1. The concept of Massed Muddler Intelligence (MMI) entails a new approach to scaling AI, emphasizing the importance of agents, local trial-and-error, and muddling through over monolithic, deterministic training models.
  2. MMIs aim to leverage the principles of embodiment, boundary intelligence, temporality, and personhood to design scalable AI systems that resemble Service-Oriented Architecture in computing.
  3. Building MMIs involves compositing different elements deliberately to create a language of differentiated forms, akin to how reinforced concrete combines materials in defined geometries to achieve specific properties.

Agent Sources

Perspective Agents • 9 implied HN points • 30 Nov 23
🕹 Technology Agents
  1. Agents in AI are rapidly evolving and have the potential to change how we access and interact with information.
  2. The emergence of personalized AI agents like GPTs can provide customized, insightful perspectives in contrast to traditional media sources.
  3. Creating trust markets with GPTs as virtual experts offers a new way to monetize expertise and deliver high-quality information.

How AI agents can improve consultations

techandsocialcohesion • 0 implied HN points • 29 Apr 24
🕹 Technology Agents
  1. AI agents can help improve consultation processes by clarifying thoughts and improving deliberation
  2. AI-powered agents could act as aides for participants, providing context, facts, and coaching to organize ideas effectively
  3. Using AI agents in consultation processes will require a balance between benefits and risks, including ensuring integrity, accuracy, and transparency

Dramatic interfaces

31 Seconds • 0 implied HN points • 04 Jan 24
🕹 Technology Agents
  1. In certain fields like government and healthcare, dealing with conflict is unavoidable.
  2. There's a need for software interfaces that can handle drama and conflict, not just transactional interactions.
  3. Dramatic intelligence is crucial for the future development of software interfaces, especially with the rise of advanced language models.