Best practice 1
Match the method to the task; do not default to an agent or a frontier LLM¶
Practice¶
- First decide whether the task at hand requires an agentAgentA system built around a language model that carries out multi-step work: it plans, calls tools, reads the results, and decides what to do next. The model supplies the reasoning; the surrounding harness does the acting. Distinct from a chat model that answers one prompt at a time..
- Choose the method that fits the task; an agent or a frontier language model is one option among several.For
- Building Effective AI Agents (Anthropic, 2024) find the simplest solution; increase complexity only when needed
- The Role of AI in Scientific Research (JRC143482, 2025) AI as a tool in science
- The GenAI Divide (2025) §4.3, pg. 13
- Expectation–Realisation Gap for Agentic AI (2026) Heterogeneity as the default
AgainstNone recorded - Prefer scripts, pipelines, workflows, and smaller models (often cheaper, faster, more maintainable).For
- AI Agents That Matter (Kapoor et al., 2024) cost-controlled evaluation; complex agents do not beat simple baselines
- The GenAI Divide (2025) Executive Summary, pg. 3
- Fine-Tuned 'Small' LLMs Still Outperform Zero-Shot Generative Models in Text Classification (Bucher & Martini, 2024) fine-tuned smaller models beat zero-shot frontier models
- EU Expert Forum on Frontier AI (2026) §3.2.3, p. 21; §3.2.2, p. 20
Qualifies- EU Expert Forum on Frontier AI (2026) §3.2.2, p. 20; ch. 4, p. 22
AgainstNone recorded - Agentic autonomy earns its cost when the task needs flexible, model-driven decisions at scale.For
- Building Effective AI Agents (Anthropic, 2024) workflows for well-defined tasks; agents when flexibility needed at scale
- Expectation–Realisation Gap for Agentic AI (2026) software engineering copilots; METR RCT
AgainstNone recorded - Add complexity only when a simpler approach falls short.
- Where a specialized method works, using a frontier model introduces avoidable failure modes.For
- Fabricated citations from LLMs (Walters & Wilder 2023) 55% (GPT-3.5) / 18% (GPT-4) of generated citations fabricated
- Leakage and the Reproducibility Crisis in ML-based Science (Kapoor & Narayanan, 2023) leakage across 294 papers in 17 fields; over-optimistic, non-reproducible
AgainstNone recorded
Before reaching for an agent, ask whether a script, a fixed workflow, or an existing specialized model already does the job as well or better. Keep the frontier model for the open-ended, judgement-heavy steps where its flexibility is the point. A method that fits the task is usually cheapest to run and easiest to reproduce.
Expose deterministic operations and specialized models as first-class tools. Incentivise the use of classical workflows where they are well established and functional. Do not assume every capability should be reached through a general model; the same demand-and-importance logic that governs where to invest in agent interfaces (BP02) applies to method choice.
Fund method choice and evaluation, not model spend. "Total token spend" is not a useful productivity metric. The goal is the cheapest reliable method that meets the need; defaulting to frontier models raises cost and reproducibility risk without a matching gain. Support the evaluation that tells adopters when a simpler method suffices.
Reasons¶
An agent is a general instrument, and general instruments are rarely the most efficient way to do a fixed job. Vendor engineering guidance is explicit: start with the simplest approach, and add agentic complexity only when a task needs flexible, model-driven decisions at scale. Agents trade latency and cost for that flexibility. When real cost is measured, complex agents often fail to beat simple baselines. On well-specified tasks, a specialized or fine-tuned model can be more accurate than a larger general model, at a fraction of the cost. The reverse choice has a record of concrete harm in science: a frontier model asked to supply references fabricates a large share of them, and treating a model as a scientific instrument without method rigour produces over-optimistic results that do not reproduce. You cannot pick the right method without first understanding the task. Understanding the task includes understanding the net benefit expected from it. Controlled trials show forecast gains from agentic tools are frequently miscalibrated and vary by task and user, so state the expected benefit against a baseline with human oversight cost deducted, then measure the realised outcome against it.
Examples¶
- A team needs to extract one field from a fixed report format every week; a short parser does it deterministically, and the agent that was drafted for the job would have cost more and varied run to run.
- A group classifying thousands of records fine-tunes a small model on labelled examples; it beats a zero-shot frontier model on their data and runs on their own hardware.
- A literature step genuinely needs open-ended judgement across messy sources, so an agent is the right tool; the routine lookups it depends on are exposed as deterministic tools it calls rather than reasons about.
- A scientist asks a chat model for supporting references and pastes them into a manuscript; several do not exist, a failure that a bibliographic search against a real index would not have produced.
- A scientist uses a literature agent that is validated against a real index to autonomously find evidence for their opinion piece. The cited papers really exist, but on review, it becomes clear that the agent only read the abstracts of the papers, citing works for sentences they do not really support. The opinion piece is rejected.
- A pipeline reports strong accuracy that collapses on re-run because the model was applied where the evaluation leaked; a simpler method with proper train/test separation would have been both honest and reproducible.
- A group expects a large speedup from an agent and, before committing, writes down the expected net benefit against its current baseline with review and correction time subtracted; for its experienced users the projected gain is marginal, so it keeps the existing workflow and re-checks the estimate on a small pilot.
Sources¶
- Anthropic, Building Effective AI Agents (2024). Prefer the simplest solution; use workflows for well-defined tasks and agents only when flexibility is needed at scale; agents trade latency and cost for performance. Vendor-primary grounding.
- Kapoor et al., AI Agents That Matter (2024). Under cost-controlled evaluation, state-of-the-art agents are needlessly complex and simple baselines can match them.
- Bucher & Martini, Fine-Tuned "Small" LLMs Still Outperform Zero-Shot Generative Models (2024). A specialized smaller model can beat a general frontier model on a well-defined task.
- Kapoor & Narayanan, Leakage and the reproducibility crisis in ML-based science (2023). Misapplied models produce over-optimistic, non-reproducible results across many fields.
- Walters & Wilder, Fabrication and errors in ChatGPT citations (2023). A concrete failure mode of using a frontier model where a retrieval method fits.
- JRC, The Role of AI in Scientific Research (2025). AI framed as a tool in science, a chosen instrument rather than a default.
- EU Expert Forum on Frontier AI (2026). Domain-specific systems can outperform general-purpose models within their domains at far lower cost (§3.2.3), qualified by the report's caution that demand for smaller models is uncertain and that non-frontier access is an imperfect substitute for frontier access.
- MIT Project NANDA, The GenAI Divide (2025). Enterprise field evidence that task type, not model capability, decides fit: users prefer AI for simple work but humans by 9-to-1 for complex or long-term work (§4.3, pg. 13). Out-of-domain business report; cited as external context.
- Lobentanzer, Expectation–Realisation Gap for Agentic AI (2026). Review of controlled trials finding that forecast gains from agentic tools are miscalibrated and heterogeneous: constrained tasks sped up while expert high-context work was slowed, so method fit is decided per task and per user rather than by an average gain. Downweighted external context; the author is a contributor to this record.
Change history¶
- 2026-07-27: Added the Expectation–Realisation Gap review (Lobentanzer 2026) as downweighted context on heterogeneous, miscalibrated benefit, with a Reasons and Examples note that method choice includes stating and re-measuring expected net benefit with oversight cost deducted (bp1-a1, bp1-a3).
- 2026-07-27: Added The GenAI Divide (MIT NANDA 2025) as a supporting source on task-based method fit (bp1-a1).
- 2026-07-27: Created as BP01. Method-selection practice at the general end of the record; grounds the disciplined stance that an agent or frontier model is chosen, not assumed. Existing BP01 through BP09 renumbered to BP02 through BP10.