Key points
- The author gathered about 800 comments from 2016 to 2026 predicting future AI limits.
- Over 100,000 votes were cast by nearly 9,700 people on the subset of claims presented.
- Roughly 8 percent of the statements focused on whether AI passes the Turing test.
A Lobsters AI tag post published in late October 2026 reviews whether artificial intelligence has met specific capability thresholds predicted by online commenters over the previous decade. According to the source material, The author collected approximately 800 comments posted between 2016 and 2026 that used phrases like “wake me up when AI can” or I'll believe it when.reportedA recent online post checks if AI has finally done the things skeptics said it couldn't do in the past.A 2026 retrospective evaluates 800 historical capability predictions against current model performance via community voting. These statements represented specific bars set for AI development at various points in time.
The claim
The post presented a selection of these historical quotes alongside voting results indicating whether the stated goal had been achieved. The voting period closed on the Friday following the post's appearance on the Hacker News front page on October 1st. By that time, more than 100,000 votes had been recorded from almost 9,700 unique participants. The author noted that using a model for a few days creates an intuition for its capabilities, which naturally raises expectations that would have seemed impossible earlier.
The response
The author highlighted the Turing test as a primary example of moving goalposts. While a plain version of the test receives a clear affirmative vote, adding qualifiers reduces the pass rate. For instance, a July 2020 comment predicting an arms race for human-level coherent articles was voted on as an AI writing such content or passing the test. Another December 2021 comment regarding auto content generators passing a restricted Turing test also received votes.
What the evidence means
The analysis suggests that simple benchmarks are often cleared quickly, but nuanced qualifications make success harder to claim. A March 2016 comment stating a chatbot passing the Turing test would happen at the very end was evaluated. An August 2017 comment describing a Turing test pass as society-shaking was also reviewed. The author emphasized that without keeping notes or a blog, it is difficult for daily LLM users to recall what would have impressed them in previous years.
In practice
Enterprise IT buyers can use this retrospective to contextualize current model limitations. The source indicates that about 8 percent of the reviewed statements concerned the Turing test, showing where consensus exists and where it fractures. The voting results for the cherry-picked quotes provide a community gauge of progress. However, the material does not confirm that all 800 collected comments were voted on, only that a subset was presented with results.reported This limits the statistical generalizability of the findings.
Why it matters
- The high vote count offers a large sample of community sentiment on AI capability milestones.
- The Turing test results demonstrate how adding qualifiers changes the perceived success of AI systems.
- The collection spans a decade, providing a long-term view of shifting expectations in AI development.
- The lack of voting on all 800 comments means the results represent a curated subset, not the entire dataset.
Topics: AI capabilities, community review, Turing test, Lobsters, Hacker News
Updated
Then / now / next
Commenters from 2016 to 2026 set specific capability bars for AI using skeptical language.
A 2026 review allows a community to vote on whether those specific bars have been cleared.
Observers should watch how future qualifiers on these tests affect the perception of AI milestones.
Story timeline
- Then Commenters from 2016 to 2026 set specific capability bars for AI using skeptical language.
- 11 Nov 2026 Wake me up when AI can
- Now A 2026 review allows a community to vote on whether those specific bars have been cleared.
- Next Observers should watch how future qualifiers on these tests affect the perception of AI milestones.
The strongest doubt
Counterpoint from the sources
The author admits that the presented quotes are cherry-picked, meaning the voting results do not reflect the outcome of the entire 800-comment collection.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Terms in this story
Turing test
A test of a machine's ability to exhibit intelligent behavior equivalent to a human.
Lobsters
A community-driven link aggregator with a focus on technology and science.
Hacker News
A social news website focusing on computer science and entrepreneurship.
Cherry-picked
Selected carefully from a larger set to support a particular view or outcome.
Questions readers ask
How many comments were included in the review?
The author collected about 800 comments from 2016 to 2026, but only a cherry-picked subset was presented for voting.
What was the voter turnout?
Over 100,000 votes were cast by nearly 9,700 people on the Hacker News front page.
Did the author vote on all collected comments?
The material states the author provided cherry-picked quotes with results, implying not all 800 comments were evaluated.
Connected stories
Reading receipt
- The author gathered about 800 comments from 2016 to 2026 predicting future AI limits.
- Over 100,000 votes were cast by nearly 9,700 people on the subset of claims presented.
- Roughly 8 percent of the statements focused on whether AI passes the Turing test.