At this point so much of what I'm reading is getting repetitive. That's good; it means we're defining some of the boundaries and pinning down attributes and trying to assimilate this tool into an existing structure. That doesn't mean it'll work, but when many people independently come to the same conclusion or generate similar ideas, there's probably some truth to those conclusions and ideas.
In the last couple of weeks, a few colleagues have finally sat down to play with ChatGPT. I am noticing a pattern. to their reactions. First, the existential crisis I've been carrying around for several months (about a year, thinking back to my experiments with HyperWrite). They question our line of work and whether it's worthwhile. They're fascinated by the prospects. They run their very specific assignment instructions through, getting back a response that is pretty passable-- and cringe. (That lines up with Machine Learning-- the more specific the prompt, the better the results.)
This blog project has been a way for me to methodically read and record findings, taking a snapshot each week of a few different readings that are floating around. I haven't just chosen the most current; I've tried to dive into one strain about GAI and writer-bots, and explored pieces from last year or longer. With only a couple of weeks to go, I have become more comfortable with the terminology and the concepts, but the presence and applications of GAI...I'm still skeptical about how good it really is for us to unleash, about how it will really make us more productive, or better at learning. And I am certainly not the only one.
"Why humans will never understand AI," David Beer for the BBC, April 7, 2023
The "Machine Minds" series within the BBC site has a lot to ponder. The black box nature of certain AIs right now still really bothers me, but more than that, trying to get my head around AI and its capabilities and inner workings makes me think of the old ghost int he machine. Or a sinister mirror-world. Or a labyrinth of labyrinths. Or...
"Will ChatGPT Change How Professors Assess Learning?" Beckie Supiano for The Chronicle of Higher Education, April 5, 2023
This piece highlights the concatenated nature of too many subsequent demands on higher ed. Rethinking assessment in the age of too many bubble sheets. Then rethinking teaching in the age of being all online. Now rethinking meaningful activities and assessments with the onset of GAI. Professors NEED support-- financial and otherwise.
"AI is flooding the workplace, and workers love it," Rani Molla for Vox, April 10, 2023
More than anything, people love technology that replaces tedium. This piece spotlights a few uses of that. Of course, it seems like most modern jobs are based around maintaining technology that has replaced other functions. Maintenance doesn't grab the spotlight the way coming up with and introducing splashy new ideas does, or the excitement of reaching milestones. But shirk on maintenance at your peril.
"Generative AI Startups in 2023," Shelby Hiter for eWeek, April 5, 2023
Obviously there's a gold rush underway, and we'll see plenty of bling and dross over the next several months or years. The list covers all kinds of GAI-- not just for writing and images, but even pharmaceuticals and the patent process.
"The best AI chatbots: ChatGPT and other fun alternatives to try," Sabrina Ortiz for ZDNet, April 12, 2023
If you can't beat them, join them, I guess.
ความคิดเห็น