# Goalposts: Vote on HN's AI Challenges

> Published 2026-10-01 · https://www.promptzone.com/wayan_chakraborty/goalposts-vote-on-hns-ai-challenges-1ii1

A new site called Goalposts appeared on Hacker News this week, letting users vote on which past HN-predicted AI challenges have actually been solved. The post collected 14 points and 8 comments in its first day.

The project pulls historical HN threads that posed specific AI milestones and turns them into simple yes/no votes. Visitors see the original claim, current status, and community tallies without needing an account.

> **Site:** Goalposts | **Votes cast:** public tallies | **Source thread:** 14 points, 8 comments | **Link:** [stoppels.ch/goalposts](https://stoppels.ch/goalposts/)

## How Goalposts Works

Users browse a list of past HN challenges such as "AI that can code at junior developer level" or "real-time translation without cloud calls." Each entry shows the original post date and a vote count.

Voting is one-click. Results update live and stay visible to everyone. No login or wallet is required, keeping participation low-friction.

## Current Numbers and Activity

The launch thread sits at 14 points with 8 comments. Early votes cluster around language-model capabilities that have clearly arrived and robotics tasks that remain unmet.

Commenters noted that several 2018–2020 predictions on image generation accuracy have crossed the threshold, while autonomous driving timelines still lag.

## How to Try It

Visit [stoppels.ch/goalposts](https://stoppels.ch/goalposts/) directly in any browser. Scroll through the listed challenges and click the vote buttons that match your assessment. Results refresh without page reload.

No installation or API key is needed. The site works on mobile and desktop.

## Pros and Cons

- Simple interface requires no setup
- Directly references verifiable HN threads
- Public tallies reduce single-author bias
- Limited to challenges that appeared on Hacker News
- No verification step for vote quality
- Small sample size so far (under 50 total votes visible)

## Alternatives and Comparisons

Other AI progress trackers include the Epoch AI dashboard and the Metaculus prediction platform. Goalposts differs by anchoring strictly to Hacker News claims rather than expert forecasts or academic papers.

| Tracker       | Source base     | Voting required | Update speed |
|---------------|-----------------|-----------------|--------------|
| Goalposts     | HN threads      | Yes             | Live         |
| Epoch AI      | Papers & data   | No              | Monthly      |
| Metaculus     | Community questions | Yes        | Daily        |

## Who Should Use This

Researchers tracking community sentiment on AI timelines will find the direct HN linkage useful. Developers deciding which capabilities to treat as solved versus still experimental can scan the vote distribution quickly.

Skip it if you need peer-reviewed benchmarks or detailed capability breakdowns; the site only records binary votes.

## Bottom Line / Verdict

Goalposts turns scattered HN predictions into a living, vote-driven scoreboard of AI progress.

The approach surfaces which milestones the community now considers crossed and which still feel distant, giving practitioners a fast reality check against earlier hype.