When two people have guns and want so very badly to shoot the other, one of those people putting down their gun and begging for mercy is likely to end very poorly for that person. You’re effectively saying Democrats should put down their guns and beg for mercy. How well do you think that’ll end when Republicans, including the conservatives on the Supreme Court, see that begging as an excuse to open fire?
You are assuming that the Do Nothing Democrats are holding real guns in the first place. They are street performers who only use props. They are perfectly happy "fighting for" issue x, y, and z, because that battle gives them a spotlight and a monologue to beg for donations. Actually solving the issue would mean taking their final bow and facing the risky, exhausting work of developing a whole new production that may not be as successful. They'd rather lose the war heroically to keep the show alive than win quietly and have to go home.
Levie's point is spot on, but I would refine it slightly. The further a CEO sits from the people actually doing the work, the more opaque the details of that work become. AI just makes this worse. Its output, on initial inspection, appears "finished", which gives a CEO, who is already well insulated from the real work being done, one more reason not to look for it.
In my experience, I spend considerably more time reviewing AI output than I would reviewing the same work from a human. With a human I need to focus on the logic, coherence, and accuracy of their work. With AI, I get a polished result, but I have to dig deep into the topic to evaluate the veracity of its work - verify assertions, confirm it has done what i want rather than what I asked, validate references etc. To reuse a phrase from a piece I recently wrote - I need to handle the consequences of "Pattern recognition dressed in the language of expertise".
Anyone who treats AI output as a finished product rather than a first draft is skipping this essential oversight step entirely. Executives who actively strip away essential controls are neglecting their role in corporate governance.
I've been writing about what happens to organizational judgment when oversight over technology degrades gradually rather than all at once at theyellowfox.net if it's of interest.
Maybe not such an accident? They have been brow-beaten, bullied, and otherwise treated worse than dogs.... I know I would be tempted to have an oopsie if I were treated the same.
Sorry mate, I got all caught up remembering usenet and gopher that I didn't read carefully enough. Add in references to Sun SPARC labs and loading slackware from I don't remember how many floppies and I will slip into full on nostalgia mode.
I guess what I was looking for were specific learning points from your implementation, since I am well into trying something similar with my wine cellar. You can find dissertations about vibe coding, but real, practical lessons requires a ton of digging.
Sorry mate, I got all caught up remembering usenet and gopher that I didn't read carefully enough. Add in references to Sun SPARC labs and loading slackware from I don't remember how many floppies and I will slip into full on nostalgia mode.
I guess what I was looking for were specific learning points from your implementation, since I am well into trying something similar with my wine cellar. You can find dissertations about vibe coding, but real, practical lessons requires a ton of digging.
Mike, don't argue, use the open source philosophy and show your work. Share your experience of how AI can be democratizing if used correctly. That journey would be a powerful lesson in how to make the tool work for you rather than the other way around.
To the others out there, AI is just another tool at your disposal. Used judiciously, it can enhance your capabilities, relieve you of tedious activities, and speed up your work. Keep your switching costs low and you can move on when your preferred tool goes the enshittification route. Used blindly and it can make you look like a fool, lead you down dead-ends, and lock you into an ecosystem that is difficult to impossible to entangle later on. I, for one, have no problem embracing a new tool so long as I don't get locked into using it.
I have been experimenting quite extensively with AI recently and auditability of the results is probably the most fundamental problem I see with current AI models.
Because of the way the dataset has been selected and curated, it is possible to train LLMs on fully open data, which leads to auditable models.
How does this data set contribute to auditability or is this auditability in another sense that what I am thinking?
Techdirt has not posted any stories submitted by The Yellow Fox.
Refining governance
Levie's point is spot on, but I would refine it slightly. The further a CEO sits from the people actually doing the work, the more opaque the details of that work become. AI just makes this worse. Its output, on initial inspection, appears "finished", which gives a CEO, who is already well insulated from the real work being done, one more reason not to look for it. In my experience, I spend considerably more time reviewing AI output than I would reviewing the same work from a human. With a human I need to focus on the logic, coherence, and accuracy of their work. With AI, I get a polished result, but I have to dig deep into the topic to evaluate the veracity of its work - verify assertions, confirm it has done what i want rather than what I asked, validate references etc. To reuse a phrase from a piece I recently wrote - I need to handle the consequences of "Pattern recognition dressed in the language of expertise". Anyone who treats AI output as a finished product rather than a first draft is skipping this essential oversight step entirely. Executives who actively strip away essential controls are neglecting their role in corporate governance. I've been writing about what happens to organizational judgment when oversight over technology degrades gradually rather than all at once at theyellowfox.net if it's of interest.
Evidence? Oh how quaint ;)
From a slightly different perspective
Maybe not such an accident? They have been brow-beaten, bullied, and otherwise treated worse than dogs.... I know I would be tempted to have an oopsie if I were treated the same.
Oh shit
That's my wife's pin written out in seashells....
Sorry mate, I got all caught up remembering usenet and gopher that I didn't read carefully enough. Add in references to Sun SPARC labs and loading slackware from I don't remember how many floppies and I will slip into full on nostalgia mode. I guess what I was looking for were specific learning points from your implementation, since I am well into trying something similar with my wine cellar. You can find dissertations about vibe coding, but real, practical lessons requires a ton of digging.
Sorry mate, I got all caught up remembering usenet and gopher that I didn't read carefully enough. Add in references to Sun SPARC labs and loading slackware from I don't remember how many floppies and I will slip into full on nostalgia mode. I guess what I was looking for were specific learning points from your implementation, since I am well into trying something similar with my wine cellar. You can find dissertations about vibe coding, but real, practical lessons requires a ton of digging.
Mike, don't argue, use the open source philosophy and show your work. Share your experience of how AI can be democratizing if used correctly. That journey would be a powerful lesson in how to make the tool work for you rather than the other way around. To the others out there, AI is just another tool at your disposal. Used judiciously, it can enhance your capabilities, relieve you of tedious activities, and speed up your work. Keep your switching costs low and you can move on when your preferred tool goes the enshittification route. Used blindly and it can make you look like a fool, lead you down dead-ends, and lock you into an ecosystem that is difficult to impossible to entangle later on. I, for one, have no problem embracing a new tool so long as I don't get locked into using it.
Auditable?
I have been experimenting quite extensively with AI recently and auditability of the results is probably the most fundamental problem I see with current AI models.
How does this data set contribute to auditability or is this auditability in another sense that what I am thinking?