“AI 预言家”与“AI CEO”的未来蓝图
这几天，两则关于AI的新闻让我格外在意。一个像是来自不远的未来，一个则发生在当下现实，但他们所传达的观点惊人一致 - AI不再是某种“试试看”的新鲜工具，而是企业运转和社会升级的基础设施。
前OpenAI研究员 Daniel Kokotajlo 发布了一篇长文《AI 2027》，用逐月时间轴推演AI在未来两年内如何进化为超级智能，甚至触发所谓的“智能爆炸”；与此同时，Shopify CEO Tobias Lütke 则在内部发布了一份员工通告，明确提出使用AI将成为公司全体员工的“基本工作要求”。
Kokotajlo 在《AI 2027》中描绘了一条明确的加速路线图：AI将在2027年中超过人类智能，2028年初达到超级智能状态，并由此开启一场不可逆的“智能爆炸”。他的这套理论并非脱离现实空谈。三年前，他写下的《2026年会是什么样》已经准确预测了多个AI发展节点，被不少AI安全领域的专家视为“可信”的预警。
这篇报告最有价值之处，不在于它一定会实现，而是它迫使我们重新设想自己的AI时间表：我们距离那场根本性转变，还有多远？
相比之下，Lütke 的操作直接得多。他在今年3月发布内部通告：从现在起，Shopify的每位员工都必须掌握并使用AI。尤其引发争议的一项公司新规定是，任何团队若要招聘，必须先证明这项工作是否可以通过AI先完成。
也就是说，AI是第一个被考虑的“员工”，人类则排在了后面。
这不仅是一个文化上的转变，更是制度层面的变革。Shopify甚至将员工“是否能够有效使用AI”纳入了表现评估。未来，一个设计师不是只交出作品就完事了，而是要展示他如何用AI辅助设计、优化流程、缩短时间。Lütke把AI比作互联网或电力，它就如同是基础设施，不可或缺。这不是“可以试试”，而是“不用就落后”。
当然，这两则新闻立刻在科技圈引发了截然不同的声音。面对《AI 2027》，不少技术社区的人直接反驳，认为这是“AI科幻小说”；甚至有人讽刺说“连上帝都跳不过官僚体制”，更别说AI了。但无论赞同或反对，这份报告的讨论度之高，足以证明它击中了许多人内心深处的不安。
而Shopify的通告虽然在商界获得不少赞扬，被认为是“迈向AI未来的范本”，但程序员群体则反应冷淡。他们担心这类政策会变成“管理层的AI打卡KPI游戏”，评估标准变得形式主义，只看“你用了多少AI”而不是“你是否创造了价值”
如果说《AI 2027》画的是一幅未来蓝图，那么Lütke则已经开始在现实中“画线”。二者虽出发点不同，但传递出的信号是一致的：AI革命的速度，远比我们想象中更快。
你可以认为《AI 2027》是危言耸听，也可以质疑Shopify是不是太激进，但你无法否认一个事实：那些已经在组织中大规模部署AI的公司，正在抢占先机。一个企业是否能“AI化”，将很快决定它的市场竞争力，甚至是它的生存能力。
因此，当我们还在犹豫要不要让AI参与工作时，有些公司已经在用AI替代人力；当我们还在调试代码、撰写文案、设计封面时，AI也许已经能自动完成80%的任务。现在的重要问题已经不再是“AI是否会改变世界”，而是“我们能不能跟上变化的节奏”。这已经是一场认知与行动的竞赛。
愿我们在这场变革中，不只是围观者，而是有意识的参与者。
Over the past few days, two pieces of AI-related news have really caught my attention. One feels like a glimpse into the near future, the other is grounded in the present. But both send the same unmistakable message: AI is no longer a novelty or something to experiment with. It’s becoming the foundational infrastructure of how businesses operate and how societies evolve.
Former OpenAI researcher Daniel Kokotajlo recently published a long article titled AI 2027, in which he lays out a month-by-month roadmap predicting that AI will evolve into superintelligence within the next two years, potentially triggering what he calls an “intelligence explosion.” At the same time, Shopify CEO Tobias Lütke issued a company-wide internal memo, declaring that effective use of AI is now a “baseline expectation” for every employee.
In AI 2027, Kokotajlo paints a detailed acceleration timeline: AI will surpass human intelligence by mid-2027 and reach superintelligent capabilities by early 2028, marking the beginning of a radical transformation. And this is not just speculative fiction. Three years ago, Kokotajlo published a post titled What 2026 Will Look Like, which accurately predicted several key developments in AI. Many experts in the AI community now view him as a credible voice of the trend.
What makes this report valuable is not that it’s guaranteed to be right, but that it forces us to rethink our assumptions about AI timelines. Just how close are we to a fundamental, irreversible AI reform?
Lütke’s move, on the other hand, is far more immediate and pragmatic. In March, he announced that every employee at Shopify must not only be familiar with AI but actively use it as part of their job. Perhaps the most controversial part of this policy? Any team that wants to hire new staff must first prove that the task can’t be completed with AI.
In other words, AI has become the first employee considered. Human hires come second.
This marks not only a cultural shift, but a structural one. Shopify has even made AI proficiency a key competency in performance reviews. It’s no longer enough for a designer to simply deliver a polished project, they must also show how AI was used to assist, streamline, or speed up the process. As Lütke put it, AI is now as essential as electricity or the internet. This is no longer a “nice-to-have.” It’s a “use it or fall behind” kind of moment.
Naturally, these two developments sparked intense reactions in tech circles. AI 2027 drew strong pushback from some online communities, who dismissed it as speculative “AI sci-fi.” One popular comment stated, “Even God can’t get past bureaucracy, so how would AI?” Still, the amount of discussion generated shows how deeply it resonated with people’s anxieties about AI’s pace.
Meanwhile, Lütke’s memo was widely praised in the business world as a bold blueprint for the future of AI adoption. But within developer communities, reactions were far more cautious. Many feared the policy would become little more than a “check-the-box AI quota,” measuring how much AI someone used instead of whether they delivered real value.
If AI 2027 outlines a blueprint for the future, Lütke is already drawing boundaries in the present. The two may start from different places, but the message is the same: the pace of this AI revolution is faster than most of us are prepared for.
You may dismiss AI 2027 as alarmist. You may think Shopify’s approach is overly aggressive. But you can’t ignore the reality that the companies actively deploying AI today are already gaining a competitive edge. Whether or not an organization can successfully “AI-enable” itself may soon determine not just its growth, but its survival.
So while some of us are still debating whether to let AI into our workflow, other companies are already using it to replace human labor. While we’re debugging lines of code, writing ad copy, or designing layouts from scratch, AI might already be automating 80% of those tasks. The key question is no longer if AI will change our world, but whether we’re able to keep up with the pace of change.
This is no longer just a tech story. It’s a race between awareness and action.
In this era of transformation, may we not stand by as mere spectators but step forward as conscious, intentional participants.