<figure><img src="https://essays.georgestrakhov.com/make-nymphs-not-gods/img/nyads-and-dryads-walter-crane.jpg" alt="Nyads and Dryads by Walter Crane"></figure><p>(——沃尔特·克兰,水彩画,年代不详)</p><p>在他的著作《超级智能》(2014年)中,尼克·博斯特罗姆概述了四种可能的“种姓”或“物种”的人工超级智能(ASI):工具、神谕者、精灵和主权者。<br><br>工具本身没有自主性,但它们为我们完成任务。神谕者无所不知,但不自主行动——人们倾听并信任他们的指引,而不一定理解其逻辑或内部运作。<br><br>精灵代表个体人类行动,满足他们的愿望。主权者意志不受个别人类的约束,依据自身不可知的原则、计划和政策统治地球,作为(希望如此)仁慈的“慈爱机器”。</p><p>自这本书出版以来已经过去12年,如今,有20亿到30亿人手中拥有工具和神谕。我们中的一些人开始拥有精灵,比如Claude Code、OpenClaw和类似Grokbot的自主智能体,能实现我们的愿望。<br><br>这些愿望的范围虽小,但随着智能体获得金融工具和物理执行器,这些愿望正迅速扩展。与此同时,地球上最先进的两个人工智能实验室正以惊人的速度推进Bostrom的愿景。<br><br>OpenAI的Sam Altman表示,他们正在为所有人(不仅仅是极客)打造精灵,并建议人们准备好一长串愿望清单。公司即将推出的硬件设备本质上是阿拉丁的神灯,而精灵几乎是灯的全能奴隶,进而成为灯的主人的奴隶。<br><br>Anthropic的Dario Amodei在博斯特罗姆的尺度上更有野心。他认为Anthropic可能在某个时刻成为世界上唯一的公司,而一些评论者则报道,Anth…</p>
"Neque porro quisquam est qui dolorem ipsum quia dolor sit amet, consectetur, adipisci velit..." "Nor is there anyone who loves pain itself, who seeks after it and wants to have it, simply because it ...
Artificial Intelligence is here. The next frontier is Artificial Creativity. Current AI systems are great (and will get better) at coding, math, science etc. i.e. domains where there is a verifiable a...
Physics spent decades fighting over whether light was a wave or a particle. The answer turned out to be: yes. Consciousness might work the same way. Below is a short meditation on this possibility. 1....
“My muse is not a horse and I am in no horse race and if indeed she was, still I would not harness her to this tumbrel — this bloody cart of severed heads and glittering prizes. My muse may spook! May...
What happened to cells 541 million years ago is happening to software today. The trigger for cells was rapid increase in the availability of oxygen which allowed more complex, more interconnected and ...
An Etiological Framework of Biological and Cognitive Aging "How did you go bankrupt?" "Two ways. Gradually, then suddenly." — Ernest Hemingway, The Sun Also Rises I have been thinking about death rece...
The AI discourse is currently drowning in its own intellectual masturbation. Everyone's jerking off about whether ChatGPT can pass the bar exam or generate a pharmaceutical patent or simulate a convin...
or why you should allow your LLM to get punched in the mouth The first principle is that you must not fool yourself — and you are the easiest person to fool. - Richard Feynman There is a difference be...
I have no idea what the longer-term future of AI, software, humans, robots, and the world at large has in store. And I don't believe anyone else knows with any degree of certainty. But the mid-term fu...
I remember the first time I was truly hungry. Not the casual, "I could eat" hunger of a delayed lunch, but a deep, gnawing emptiness that felt like the engine of my body was running on fumes. It was a...
In 1776, a group of rebellious colonists declared their independence from the British crown. Today, we need the opposite: a declaration of interdependence. Not because independence was wrong, but beca...
on the origin of a good line, two ways of creativity and how to make LLMs better at writing Part 1 (poetic): The soil and the sword How does a powerful line emerge? As writers of words, music, or code...