I love LLMs, I hate hype
496 points
• 4 days ago
• Article
Link
作者对当前的人工智能发展充满热情，并以自己长期从事该领域的职业生涯为凭证。从新型语言模型、自动驾驶汽车，到视频生成和高级编程代理，进步触手可及。作者最近在本地编程工具上的一些实验甚至让他半开玩笑地宣称，期待已久的 Linux 桌面元年终于来临，因为像配置复杂软件这样的任务，现在只需发出一条自然语言命令就能完成。
尽管如此，他强烈反对围绕这个行业的负面炒作。作者特别抨击那些声称人们正无可救药地落后、或机会窗口正在迅速关闭的论调，认为这种说法具有操纵性，旨在让个人产生自卑感，并以此逼迫他们在虚假前提下迁往像 San Francisco 这样的特定科技中心。
他同样驳斥把 AI 从"高级自动补全工具"一跃想象成注定要统治整个光锥的荒诞逻辑。认为会有某种突如其来的神秘奇点，让不在特定社交圈里的人一夜之间被抛弃的想法是可笑的。相反，作者认为那些散布末日式预言的人，往往是在投射自身的不安，用关于安全或地缘政治竞争的高调论述来掩饰对控制权的渴望。
在他看来，问题的核心是对商品化的恐惧。 AI 的进步很大程度上源于摩尔定律和计算力的整体提升，而非某些前沿实验室的专有发明。这些机构有明显的经济动机去维持一种错觉——把所有突破都归功于自己，因为这种叙事可以为其高估值辩护，并确保数十亿美元的融资。通过抹黑开源贡献，他们试图对本质上不可避免的技术保持控制。
最后，作者反思了编程性质的变化，超越了最初对大型模型能力的怀疑。虽然这些工具不能取代人的推理，但它们像编译器或搜索引擎一样，成为强有力的倍增器。作者也承认，这些模型可能带来认知疲劳并常产生低质量内容，但它们代表了计算机革命中一种必然且有用的演进。 AI 不是一种从根本上打破软件开发规律的神奇颠覆，而是几十年来技术轨迹的延续。
The author expresses profound enthusiasm for the current state of artificial intelligence, citing a career deeply embedded in the field. From new language models and self-driving cars to video generation and advanced coding agents, the progress is palpable. Recent experiments with local coding tools have even led the author to jokingly declare that the long-awaited year of the Linux desktop has finally arrived, as tasks like configuring complex software become as simple as issuing a natural language command.
Despite this excitement, there is a strong rejection of the pervasive, negative hype surrounding the industry. The author is particularly critical of narratives suggesting that people are falling hopelessly behind or that a narrow window of opportunity is rapidly closing. This perspective is viewed as manipulative, designed primarily to make individuals feel inadequate and to pressure them into relocating to specific tech hubs like San Francisco under false pretenses.
Equally dismissed is the dramatic leap in logic that frames AI as a transition from a sophisticated autocomplete tool to an entity destined to dominate the entire light cone. The idea that a sudden, mysterious singularity will change everything overnight for those not involved in the right social circles is labeled as absurd. Instead, the author argues that those propagating such apocalyptic scenarios are often projecting their own internal anxieties, attempting to cloak their desire for control in high-minded arguments about safety or geopolitical competition.
The core of the issue, according to the author, is the fear of commodification. AI progress is largely the result of Moore's Law and general advancements in computing power rather than the proprietary achievements of specific frontier labs. These organizations have a clear financial incentive to maintain the illusion that they alone are responsible for these breakthroughs, as this narrative justifies their massive valuations and secures billions in funding. By discrediting open-source contributions, they attempt to maintain a grip on technology that is fundamentally inevitable.
Finally, the author reflects on the changing nature of programming, moving past initial skepticism about the capabilities of large models. While these tools do not replace human logic, they function as powerful force multipliers, similar to how compilers or search engines previously transformed the field. The author acknowledges that while these models can induce cognitive fatigue and often produce low-quality content, they represent an essential, useful evolution in the computer revolution. AI is not a magical disruption that fundamentally breaks the laws of software development, but rather a continuation of the same technical trajectory that has been unfolding for decades.
322 comments • Comments Link
在 AI 辅助维护的推动下，fork 开源项目变得愈发容易，这意味着我们正进入一个"想怎么定制就怎么定制"的时代——个性化定制往往被置于向上游提交改进之上的优先级。
开源的长期价值不仅在于代码本身，更在于共享的传统与文档；这些对于项目保持可用性至关重要，尤其是在像科学计算这样高度依赖领域知识的复杂领域。
管理分支（fork）仍然是一项挑战。尽管越来越多的 AI 工具用于追踪上游变更并解决冲突，但这也在个性化软件与长期维护负担之间带来了权衡。
许多用户更看重能满足特定需求的"够用"软件，这表明并非每个项目都需要持续更新或企业级的维护，因此主要软件套件的订阅流失对个人使用场景而言并不那么重要。
尽管高端 AI 推理的成本目前因市场争夺而被补贴，但供应商之间的良性竞争以及本地执行的选项表明，访问智能模型的成本随着时间很可能会下降，而非上升。
与 AI 模型进行头脑风暴是开发者的一条高效捷径，它能为项目提供即时的切入点，减少"停工"时间；但这同时需要批判性判断，以避免盲目采纳会导致技术债务的建议。
围绕 AI 的炒作常在末日般的恐惧与乌托邦式的承诺之间摇摆，二者都旨在推动投资并制造焦虑；应对之道是保持冷静与专注，避免陷入行业话语中常见的负面情绪。
关于是否应将大型语言模型（LLMs）称为"AI"，业界存在分歧。一些人认为该术语过于简化，可能掩盖对真正人工智能研究的合理区分；另一些人则认为，对于那些能够实现以往需要人类智能才能完成的任务的技术，称其为"AI"是有用且恰当的描述。
对 AI 生成创作的怀疑常集中于所谓缺乏"灵魂"或原创性，但这与早期对计算机生成图像的批评如出一辙，说明其质量与实用性将继续快速提升。
企业对难以实现的独角兽估值以及对控制计算能力的渴望，仍然是行业不稳定的主要动力，这往往促使短期的市场营销而非长期的可持续发展。
总体来看，这场讨论反映出 AI 作为生产力工具的不可否认的效用，与围绕该行业的系统性焦虑之间存在深刻张力。在那些把这些模型仅当作需要掌握的软件抽象的人，与担心"AI"标签掩盖缺乏根本性创新的人之间，存在明显隔阂。尽管一些专业的软件匠人通过把 AI 当成克服障碍的协作伙伴而获得成功，人们仍普遍担心"炒作"文化会被用来攫取价值并操纵职业预期。归根结底，尽管这项技术本身功能强大且正在迅速改进，但它对编程的社会与结构基础——以及开源社区可持续性——的长期影响，仍然不确定且备受争议。 • The current ease of forking open-source projects, driven by AI-assisted maintenance, suggests a shift toward a "have it your way" era where upstreaming improvements is less prioritized than individual customization.
• The long-term value of open source lies not just in the code itself, but in the shared traditions and documentation, which remain vital for projects to stay usable, especially in complex fields like scientific computing where domain knowledge is essential.
• Managing forks remains a challenge, though AI tools are increasingly used to track upstream changes and resolve conflicts, potentially leading to a trade-off between individualized software and the burden of long-term maintenance.
• Many users prioritize software that is "good enough" for their specific needs, suggesting that not every project requires constant updates or enterprise-grade maintenance, making the "subscription churn" of major software suites less relevant to individual use cases.
• While high-end AI inference costs are currently subsidized to capture market share, healthy competition among providers and local execution options suggest that the cost of accessing intelligent models will likely decrease, not increase, over time.
• Brainstorming with AI models acts as a productive bridge for developers, eliminating "downtime" by providing immediate starting points for projects, though it requires critical judgment to avoid blindly accepting suggestions that create technical debt.
• The hype surrounding AI frequently oscillates between apocalyptic fear and utopian promises, both of which serve to drive investment and create anxiety; navigating this requires intentionality to avoid the "negative valence" often found in industry discourse.
• Disagreement exists over the labeling of LLMs as "AI," with some arguing the term is reductive and obscures legitimate research into genuine artificial intelligence, while others view it as a useful descriptor for technology that achieves results previously requiring human intelligence.
• Skepticism toward AI-generated creative works often focuses on a perceived lack of "soul" or originality, though this mirrors early criticisms of computer-generated imagery, suggesting that quality and utility will continue to improve rapidly.
• Excessive corporate focus on unattainable unicorn valuations and the desire to control compute power remains a primary driver of industry instability, often incentivizing short-term marketing over sustainable development.
The conversation reflects a deep tension between the undeniable utility of AI as a productivity tool and the systemic anxieties surrounding its industry. There is a clear divide between those who view these models as just another layer of software abstraction to be mastered and those who worry that the "AI" label masks a lack of fundamental innovation. While professional software craftspersons are finding success by treating AI as a collaborative partner to bypass roadblocks, there is a shared concern that the culture of "hype" is being weaponized to extract value and manipulate career expectations. Ultimately, the discourse suggests that while the technology itself is powerful and rapidly improving, its long-term impact on the social and structural fabric of programming—and the sustainability of the open-source community—remains uncertain and highly contested.