Back
FutureSearch Research Publications
webfuturesearch.ai·futuresearch.ai/research/
FutureSearch uses AI-assisted forecasting to analyze AI development trajectories; their publications may be useful for those researching AI timelines, risk estimation, and evidence-based safety planning, though content could not be directly verified from the provided metadata.
Metadata
Importance: 42/100homepage
Summary
FutureSearch is an AI-assisted forecasting and research organization that publishes work on predicting future developments, including AI timelines and safety-relevant forecasts. Their research applies probabilistic reasoning and superforecasting techniques to questions of technological and societal change. The publications page aggregates their outputs relevant to AI risk and futures analysis.
Key Points
- •Applies superforecasting and probabilistic methods to AI timelines and safety-relevant future scenarios
- •Produces research on questions like transformative AI arrival dates and associated risks
- •Bridges forecasting methodology with AI safety considerations to inform planning and policy
- •Outputs are intended to support evidence-based reasoning about long-term AI trajectories
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| FutureSearch | Organization | 50.0 |
Cached Content Preview
HTTP 200Fetched Mar 20, 20264 KB
# EveryRow Agent
Research, then analyze
An agent researches each row on the web, then analyzes what it finds. Add columns of data you don't have yet.
[Try Research free](https://cohort.futuresearch.ai/app) [Give to your AI](https://github.com/futuresearch/futuresearch-python)
▶ 2-min demo video coming soon
1–11¢per row
[accuracy verified](https://evals.futuresearch.ai/) in [Deep Research Bench](https://evals.futuresearch.ai/)
💰
company → annual\_price, tier\_name
### Enrich SaaS Pricing
Research pricing pages for hundreds of products. Extract tier names, annual prices, and feature lists as structured columns.
$6.68 • 99.6% success • 246 products
[Tutorial →](https://futuresearch.ai/docs/add-column-web-lookup) [Try it →](https://cohort.futuresearch.ai/app)
🏷️
job\_title → category, seniority
### Classify Job Postings
Add category, seniority level, and confidence columns to job listings using LLM classification.
$1.74 • 100% success • 200 postings
[Tutorial →](https://futuresearch.ai/docs/classify-dataframe-rows-llm) [Try it →](https://cohort.futuresearch.ai/app)
📦
package → days\_since\_release, contributors
### Research Package Metadata
Look up days since last release, contributor counts, and other metrics for PyPI packages from the web.
$3.90 • 1.3¢/row • 300 packages
[Tutorial →](https://futuresearch.ai/docs/rank-by-external-metric) [Try it →](https://cohort.futuresearch.ai/app)
## Give your AI a team of agents
### Claude Code
```
claude mcp add futuresearch --scope project --transport http https://mcp.futuresearch.ai/mcp
Then ask Claude to research your data.
```
[Agent setup guide →](https://futuresearch.ai/docs/claude-code)
### Python SDK
```
pip install futuresearch
from futuresearch.ops import agent_map
result = await agent_map(
task="Find the annual price
of the lowest paid tier",
input=products_df,
response_model=PricingInfo
)
```
[Get API key →](https://cohort.futuresearch.ai/app) [Docs →](https://futuresearch.ai/docs/reference/RESEARCH)
## Pricing
Start with **$20 in free credits**. No credit card required. Pay only for what you use—costs scale with research complexity.
| Task | Rows | Cost/row | Success |
| --- | --- | --- | --- |
| [SaaS pricing lookup](https://futuresearch.ai/docs/add-column-web-lookup) | 246 | 2.7¢ | 99.6% |
| [Job classification](https://futuresearch.ai/docs/classify-dataframe-rows-llm) | 200 | 0.9¢ | 100% |
| [Package metadata](https://futuresearch.ai/docs/rank-by-external-metric) | 300 | 1.3¢ | — |
### Why costs vary
Every row gets its own web research agent. Agents have degrees of freedom. They spend more tokens doing more research for harder tasks. Simple lookups finish quickly; complex research requires multiple page visits and reasoning steps.
## Resources
### Tutorials
- [• SaaS prici
... (truncated, 4 KB total)Resource ID:
8330d6cca4a886c9 | Stable ID: YTY1NDNlY2