可研究網路各處任意主題的 AI 代理技能,並彙整出有根據的摘要。
Last30Days 是一款研究代理,能找出人們最近對任何主題的真實看法,而非演算法或編輯部認定的重要資訊。提供人物、公司、產品、技術或「X vs Y」比較等主題,它便會搜尋過去 30 天社群平台與開放網路上的討論,包括 Reddit 討論串、Hacker News 討論串、YouTube、GitHub 活動、Polymarket 預測市場等。
它不會只列出連結,而是依據真實互動熱度排序結果——包括讚數、喜歡數,以及有真金白銀支撐的市場賠率——將重複故事合併為單一群集,並整理成有根據的簡報,每項主張都附有文中引用。「最佳觀點」部分會呈現社群中最犀利、最受歡迎的回應,讓你在了解事實之餘,也能掌握討論的氛圍。
可用於競爭與市場研究、追蹤產品上市或新聞事件引發的情緒、關注快速演變的科技辯論,或快速掌握任何重視新近且經社群驗證的訊息、勝過精心包裝的新聞報導的主題。最後會提供清晰易讀的報告,內容根據過去一個月真實使用者的互動整理而成。
從英文翻譯,僅供參考。
According to the current academic consensus, when is artificial general intelligence (AGI) expected to be achieved?
There is no current academic consensus on a date for AGI. As of October 2026, a specific year—whether 2030, 2047, or later—should be treated as a forecast, not an agreed scientific conclusion.
One useful reference point is a survey of 2,778 researchers who had published in leading AI venues, conducted in October 2023 and subsequently published in the Journal of Artificial Intelligence Research. Its aggregate forecast estimated, assuming scientific progress continues undisrupted:
These are different milestones, not interchangeable definitions of AGI. And “50% by 2047” does not mean researchers agree AGI will arrive in 2047; it summarizes uncertain forecasts collected in 2023. .
There is also disagreement about the technological route. The 2025 AAAI presidential-panel survey found that 76% of respondents considered scaling up current AI approaches “unlikely” or “very unlikely” to produce AGI. That finding concerns an approach, not a deadline: it does not mean those researchers believe AGI is impossible. ; .
Bottom line: The academically defensible answer is “the timing is unknown, with substantial disagreement.” If you need a numerical reference, 2047 is the 50% date from one large researcher survey for a particularly demanding human-level-AI milestone—not the current academic consensus on AGI.
本次執行費用115 點數≈ $0.115網路搜尋 ×5GPT-6.1 Sol · 52,169 詞元
執行 AI 代理會消耗你的點數。隨時都能停止——你只需為它實際產出的內容付費。