LLM Links

Aswath Damodaran on competing against your AI clone; Ethan Mollick on students and teachers using LLMs; Neo, a household robot; The Zvi rates LLMs on Silver's technological Richter Scale

Aswath Damodaran writes,

He mentioned that he had developed a Damodaran Bot, and explained that it was an AI creation, which had read every blog post that I had ever written, watched every webcast that I had ever posted and reviewed every valuation that I had made public. Since almost everything that I have ever written or done is in the public domain, in my blog, YouTube videos and webpage, that effectively meant that my bot was better informed than I was about my own work, since its memory is perfect and mine is definitely not.

While I was not as thorough, I tried to use delphi.ai to create an AI clone of myself.

Damodaran says that as these bots or clones appear, the question will be what you can do to ensure that you have some value at the margin.

It is your job, and mine, to think of the moats that we can erect (or already have) that will make it more difficult for our bots to replace us.

At this point, my sense is that I have a great deal of tacit knowledge, meaning skills and experiences that are not documented on the Internet. This is probably true of Damodaran, also. So I do not yet feel threatened by my clone. Still, it is an interesting way to frame the issue of how the future will look.

Ethan Mollick writes,

As of eight months ago, a representative survey in the US found that 82% of undergraduates and 72% of K12 students had used AI for school. That is extraordinarily rapid adoption. Of the students using AI, 56% used it for help with writing assignments, and 45% for completing other types of schoolwork. The survey found many positive uses of AI as well, which we will return to, but, for now, let’s focus on the question of AI assistance on homework. Students don’t always see getting AI help as cheating (they are simply getting answers to some tricky problem or a challenging part of an essay), but many teachers do.

…This is not a new problem. ... By 2017, a majority of students were copying internet answers, rather than doing the work themselves.

Tyler Cowen links to several tweets about Neo. Kristi Hines tweets,

1X Technologies, a Norwegian startup backed by OpenAI, has unveiled NEO, a versatile humanoid robot designed to assist in a wide range of settings, from industrial environments to homes. NEO continuously learns and improves through the fusion of its AI "senses" and physical body, allowing it to understand its environment more deeply over time and interact with humans intuitively. The robot's lightweight, soft, and compliant design prioritizes safety for close interaction with humans, and it can perform complex tasks in industries like security, logistics, and manufacturing, as well as household chores.

I have speculated that robots optimized for particular settings, such as an operating room or a kitchen, will be more effective than general-purpose robots that use the human form factor. I am willing to be wrong.

Zvi Mowshowitz writes,

for the skeptic case, the AI-fizzle world, is that AI could prove to be ‘only internet big.’ In that scenario, GPT-5-level models are about as good as it gets, and they don’t enable dramatically better things than today’s GPT-4-level models. We then spend a long time getting the most out of what those have to offer.

…I think that 5-level models, given time to have their costs reduced and to be properly utilized, will inevitably be at least internet big, but only 45% of respondents agreed. I also think that 5-level models are inevitable - even if things are going to peter out, we should still have that much juice left in the tank.

He argues, and I am inclined to agree, that the LLMs are already a very important technological innovation. The thing is, as with the Internet, it takes a while for the right apps to be built and for people to learn the best uses.

substacks referenced above:@

AI and the Technological Richter Scale

@

Beat your Bot: Building your Moat against AI

@

Post-apocalyptic education