Prediction: We will have AGI by 2025-2028

Prediction: We will have AGI by 2025-2028

【免费下载链接】LifeOS⛰️ The Life Operating System — an intent engineering platform that moves you from your current state to your ideal state, in life and work.项目地址: https://gitcode.com/GitHub_Trending/pe/LifeOS

Prediction: We will have AGI by 2025-2028

Date of Prediction: March 2023

Quote:

This is why AGI is coming sooner rather than later. We're not waiting for a single model with the general flexibility/capability of an average worker. We're waiting for a single AGI system that can do that. To the human controlling it, it's the same. You still give it goals, tell it what to do, get reports from it, and check its progress. Just like a co-worker or employee. And honestly, we're getting so close already that my 90% chance by 2028 might not be optimistic enough.Why We'll Have AGI by 2025-2028

References:

• Why We'll Have AGI by 2025-2028

Status:IN PROGRESS🔄

Notes:

• This prediction works off this definition of AGI.
• Jan 12, 2025 — This prediction has been made multiple times and I'm improving my content RAG to find the earliest instance.
• Jan 12, 2025 — I am still confident in this one, and am currently putting this at 40% chance for 2025, and 50% for 2026, and 10% 2027 or beyond.

把示例拆开，每个预测块由以下七个部分构成： | 区块 | 格式 | 说明 | |------|------|------| | 标题（H2/H3 双标题） | `## Prediction: <主题>` 与 `### Prediction: <主题>` | 同一标题重复两次，H2 用于博客索引/锚点，H3 承接正文层级；主题用一句话凝练预测内容 | | 预测日期 | `Date of Prediction: <年月>` | 记录预测提出的时间，便于回测"预测是否按时兑现" | | 引用 | `Quote:` + `

…来源

` 内嵌 ``**：引用不是孤立摘录，而是"引文 + 出处链接"的 HTML 组合，这保证了每条预测引语都能追溯到原文，符合预测档案对可验证性的要求； - **Notes 的日期前缀**：`Jan 12, 2025 — …` 的写法表明 Notes 是"流水账式"的跟踪日志，记录置信度调整（40% / 50% / 10% 的概率分布）与去重改进（RAG 查找最早实例），让预测块成为持续更新的活文档，而非一次性输出。 ### 5. OUTPUT INSTRUCTIONS：输出约束 ```text # OUTPUT INSTRUCTIONS // What the output should look like: - Only output the predictions in the format described above. - Get up to 5 references for the reference section based on the input. - Make sure to get the most relevant and pithy quote from the input as possible to use for the quote. - Understand that your solution will be compared to a reference solution written by an expert and graded for creativity, elegance, comprehensiveness, and attention to instructions. - The primary reference should be used as the quote, and that should also be used as the first reference mentioned in the reference section.

输出约束共五条，可归纳为三组：

1. 格式唯一性：只输出上述格式的预测块，不做任何额外解释；
2. 内容约束：参考文献最多 5 条；Quote 必须是输入中最相关、最精炼（pithy）的原话；
3. 一致性约束：中的主引用来源，必须同时是 References 列表中的第一条——保证"引文出处"与"参考文献首位"指向同一来源，避免引用与参考脱节。

此外，模式明确提示"你的产出将与专家撰写的参考答案对比，并按创造力、优雅度、全面性、指令遵循度评分"——这是 Fabric 系模式常见的自评压力机制，用以提升输出质量。

6. INPUT：输入占位

# INPUT INPUT:

system.md以INPUT:占位符收尾，实际使用时，待分析的内容（文章全文、访谈文字稿、播客转写等）会拼接在这里，构成一次完整的模式调用。

三、在 LifeOS 中如何执行该模式

create_prediction_block在 LifeOS 中不需要调用外部 CLI，而是走"原生模式执行"（Native Pattern Execution）路径。根据 LifeOS/install/skills/Fabric/SKILL.md：

User Request → Pattern Selection → Read system.md → Apply → Return Results

具体到本模式，调用过程是：

1. 选择模式：当用户表达"把这篇内容里的预测整理成博客预测块""生成 /predictions 页面内容"等意图时，即命中create_prediction_block；
2. 加载提示词：读取Patterns/create_prediction_block/system.md的完整内容；
3. 应用模式：将# INPUT之后的占位替换为用户的待分析内容，AI 按 IDENTITY → GOAL → STEPS → OUTPUT INSTRUCTIONS 的顺序执行；
4. 返回结果：输出符合模板的预测块序列，供用户直接粘贴到 Markdown 博客的/predictions页面。

工作流的完整细节参见 LifeOS/install/skills/Fabric/Workflows/ExecutePattern.md，其中 Step 2 给出了加载模式文件的参考实现（PATTERN_PATH="$HOME/.claude/skills/Fabric/Patterns/$PATTERN_NAME/system.md"）。关于调用边界，需要注意两点：

• Fabric 的 CLI 仅在两种场景下需要：YouTube 字幕提取（fabric -y URL）与 URL 内容抓取兜底（fabric -u URL）；
• 模式名必须精确匹配，create_prediction_block不能写成createpredictionblock；不确定时可通过fabric --list核对。

四、实战：一个完整的调用示例

假设输入是一篇科技访谈，其中包含这样的论述（模拟输入）：

我们正在接近一个转折点。我认为到 2027 年，大多数知识工作者会使用个人 AI 助理处理超过一半的日常工作。这不是炒作，基础设施成本下降的速度是 10 倍级的。

将这段内容拼接到# INPUT之后运行模式，预期产出如下（依模板组织，供参考）：

## Prediction: Most knowledge workers will use personal AI assistants for half their daily work by 2027 ### Prediction: Most knowledge workers will use personal AI assistants for half their daily work by 2027 Date of Prediction: September 2026 Quote: 

We're approaching an inflection point. By 2027, most knowledge workers will use personal AI assistants for more than half of their daily work. Infrastructure costs are dropping 10x.The Future of Work Interview

【免费下载链接】LifeOS⛰️ The Life Operating System — an intent engineering platform that moves you from your current state to your ideal state, in life and work.项目地址: https://gitcode.com/GitHub_Trending/pe/LifeOS