Wolf thinks that AI labs are building what are essentially “very obedient students” — not scientific revolutionaries in any sense of the phrase. AI today isn’t incentivized to question and propose ideas that potentially go against its training data, he said, limiting it to answering known questions.
“To create an Einstein in a data center, we don’t just need a system that knows all the answers, but rather one that can ask questions nobody else has thought of or dared to ask,” Wolf said. “One that writes ‘What if everyone is wrong about this?’ when all textbooks, experts, and common knowledge suggest otherwise.” […]
“[T]he most crucial aspect of science [is] the skill to ask the right questions and to challenge even what one has learned,” Wolf said. “We don’t need an A+ [AI] student who can answer every question with general knowledge. We need a B student who sees and questions what everyone else missed.”
Clearing out the decks here of saved tabs in my browser, quote via Matt Webb from a few weeks back:
Meanwhile blogging has become small-p political again.
Slowly, slowly, the web was taken over by platforms. Your feeling of success is based on your platform’s algorithm, which may not have your interests at heart. Feeding your words to a platform is a vote for its values, whether you like it or not. And they roach-motel you by owning your audience, making you feel that it’s a good trade because you get “discovery.” (Though I know that chasing popularity is a fool’s dream.)
Writing a blog on your own site is a way to escape all of that. Plus your words build up over time. That’s unique. Nobody else values your words like you do.
Blogs are a backwater (the web itself is a backwater) but keeping one is a statement of how being online can work. Blogging as a kind of Amish performance of a better life.
Oh except that it does work as well as those other platforms, if that’s what you want, that’s the magic.
I still believe in old school “soul” blogging. Much more so than the social media cesspool.
This piece by the founder of Powder, Barthelemy Kiss, in Business Insider (archived) is a good one that echoes something I wrote about over two years ago now, that AI needs to make room as an industry for people who are not engineers and PhDs etc by training.
Working on this startup, I’ve learned that people with a liberal arts background have a major edge in our industry. The liberal arts grads we’ve hired have a creative, human-centric approach to understanding the best applications of AI in their respective fields.
They have a hunger for constant education and a desire for self-improvement, which is essential in this space. At a core level, liberal arts disciplines hone critical thinking. We need critical thinkers to challenge the results of AI. Models can hallucinate, so you need to be able to cross-reference data. […]
While AI engineers are absolutely key, it’s only through alliances between talent from liberal arts backgrounds, who can think creatively about building applications around this tech stack, that the technology will reach its full potential.
When he started at Google Labs, Johnson initially helped write prompts for AI models. As the AI boom took off, the role of prompt engineer similarly captured the attention of other English majors interested in AI.
Johnson said that prompting will change, especially as AI models improve at rewriting prompts — but a new role, which he referred to as the “AI wrangler,” has emerged.
“That’s maybe the next stage of the prompt engineer,” Johnson said.
Johnson describes the AI wrangler role as not necessarily requiring coding expertise but involving deep knowledge of the latest models and their capabilities.
For example, if someone wants to create a 30-second animated video with AI, the AI wrangler would know the best tool for that task and how to use it, Johnson said. He said the role requires a “certain level of technical sophistication” but doesn’t require knowing how to program.
Here’s a glimpse of a conceptual art project I have been working on in the background that incorporates AI on a few different levels. The working series title is ART/PROMPT because it attempts to take generative AI prompts (or rather meta hot-takes on them) and turn those into hand-lettered paintings. So the viewer looking at the painting sees the text prompt and then their own imagination is provoked into completing the scene described.
Here’s an initial image set, all of these made in Recraft. In making these, I came to exactly the same headspace I have with other explorations in the past using AI to visualize art concepts: it’s unlikely that (1) I would be easily able to paint better paintings than these myself by hand (though they would be qualitatively very different, which is still important – probably now of primary importance), and (2) that I would be able to get as good quality photos of them as easily as I can just by figuring out how to get an AI image generator to spit them out for me.
I did do one hand-painted prototype of one of these slogans IRL, but it’s not included in this sample set cause, well, I forgot to photograph it. And again, it’s just not as good as the items generated by Recraft in this set.
The opportunities in this seem endless to me, and I have many I made that aren’t in that public set.
Maybe I’m just an asshole (probably), but there’s also something really exciting about being able to use these… let’s call them meme paintings? as a way to directly confront a lot of the outrage and unhappiness people are feeling with regard to generative AI technologies:
For the most part, those reservations are not actually feelings I really personally resonate with. Though I understand well where they are coming from anyway, and want to recognize the positions, rather than simply ignore them.
I’m actually looking currently for partners with functional robot arm painting set-ups who would be willing to work together on executing some of these as special editions. Email me and let’s talk.
It’s a bit difficult to cipher out all of what it says in those apparently 4 or so pages where my name seems to be mentioned twice (assuming Google Books results are somewhat accurate). Most of these academic-style references I have found have just been regurgitating quotes from my original Newsweek piece, in many cases, quotes that were wholly or partly written in the first by ChatGPT, which I used to help write that article.
Another in this thematic series, commenting on the first, but this one also intersecting with the ART/PROMPT series, which I’ll try to post tomorrow about in more detail. This one made in Recraft.
Loosely inspired by Rauschenberg’s Erased Dekooning (the story of which I love) and the ability of AI to simply erase people and things out of history at the touch of a button. Especially anything that might be considered politically inappropriate (or ‘out of scope’ to quote Deepseek).
