Posts by elotroalex@hcommons.social
(DIR) Post #ASF1TpHlbcoIMmchNo by elotroalex@hcommons.social
2023-02-01T16:33:57Z
0 likes, 1 repeats
Funny how when I asked what our new science should be, everyone responded by naming an existing field. Also funny that no one mentioned the obvious Library Science, considering how I described this science of the future. Why don't nobody want a new science?! Aren't y'all bored with the ones we have. Let's do a little poll.
(DIR) Post #ASF1VNbFVnR550Tc4u by elotroalex@hcommons.social
2023-02-01T14:17:00Z
0 likes, 1 repeats
Also heard Chomsky's wholesale dismissal of the new trends in generative AI yesterday. He doesn't think that this new trend can help us understand anything about the world, that it is merely useful for some things like sound-to-text, but cannot contribute to "science." I'm not sure I agree. My hunch is that it can contribute to a science that does not exist yet: a cousin of linguistics, who had an affair with book history, was raised by digital humanities, and lives next to cultural studies. In this new science what the machine generates is one of the sets of evidence we use to understand the whole of culture, the relations between texts and images as they appear on the record. This new science is less interested in statements and their relationship to thought, or "real human languages" as Chomsky insists. Our new science cares more about what's in the record, and how it's parts relate to one another and itself. What shall we call this new science?
(DIR) Post #ASF1YK29E95qt8SfGy by elotroalex@hcommons.social
2023-02-01T16:39:39Z
0 likes, 0 repeats
@CitizenWald lol. Basically. Pata-physics is taken. That was bawler. If only we could have Jarry's sense of humor.
(DIR) Post #ASPczn5zv52Rjsg9NQ by elotroalex@hcommons.social
2023-02-06T16:53:19Z
0 likes, 0 repeats
Woke up thinking about social relations again. It always struck me that in traditional Marxism you must adjust the underlying structure, the base, before you see any changes in the superstructure. Social relations belong to the base, and those are determined in capitalism by, well, capital and production. Like most things though, once you accept you can't get everything by radical revolution, can't we think of some social relations that can change without a total transformation of the base? Take DH within universities. Let's look at 2 of the relevant roles here: faculty and service staff. Now we are told in DH that collaboration is key. I agree, of course. But collaborating as faculty is weird. There are no incentives, not any real ones anyway. That's maybe because the university already baked in a social relation of service in the service staff. They get paid to be the FT collaborators. What if though we could change that social relation?
(DIR) Post #ASPczncFz8qJLvlueu by elotroalex@hcommons.social
2023-02-06T16:54:39Z
0 likes, 0 repeats
To be fair, what I describe sounds like a big revolution relative to the way things can change around here. It's not even what I had in mind when I said radical revolution, which involved overthrowing capitalism. 🤣
(DIR) Post #ASPczp3wbgaHq5jXSC by elotroalex@hcommons.social
2023-02-06T18:29:35Z
0 likes, 0 repeats
@jose_eduardo I should be more specific. At the "elite" universities faculy get rewarded for books... with "good" presses. That's the reward. Solitary work. That's how you achieve Full Professor. There is no other way. At least in the humanities.
(DIR) Post #ASPczq9IZGSfCzFKRE by elotroalex@hcommons.social
2023-02-06T18:30:15Z
0 likes, 0 repeats
@jose_eduardo to be even more clear: books are a lot of work. A lot. There's hardly any time left to do anything else besides teaching and service to the department.
(DIR) Post #ASecQfQhTaONKKu78q by elotroalex@hcommons.social
2023-02-13T21:46:02Z
0 likes, 1 repeats
I just put in my resignation as one of DHQ editors after many happy years of good work. During my tenure as Internationalization Editor, we did some great work there expanding the readership and reviewer pool, plus some really awesome special issues in French, Spanish & Portuguese. No dark reasons for leaving, just time to move on to other projects and new commitments. The journal continues to be such a cornerstone of international DH. Here's wishing Julia Flanders, Melissa Terras and the rest of the team wind beneath their sails! Godspeed, friends! 💖
(DIR) Post #ATAsD3tnWFF9UJlZFQ by elotroalex@hcommons.social
2023-03-01T14:28:12Z
0 likes, 0 repeats
@TedUnderwood this personality business is what caught my attention the most when ChatGPT roled out. As soon as we started understanding a little what reinforcement learning was doing, I was like "personality!" And we'll see many many more "persons" in the foreseeable future. My studied prediction is that the personalities themselves will be eventually generated. I mean, we will be able to not only generate text based on a specifict training method, but we will be able to generate training methods based on prompts. Taking bets this is happening before I die.
(DIR) Post #ATLHJS2dmygwlPtzVI by elotroalex@hcommons.social
2023-03-06T14:56:31Z
0 likes, 0 repeats
@TedUnderwood this post had me spending some morning headaches on trying to understand why logits are so important for AI. I wonder if you and @sramsay would partner up to do a Statistics for Humanists to folllow up on Stephen's and Patrick's _Six Septembers_. I'd pay top dollar for the hard cover.
(DIR) Post #ATPLmHDuGbxw3g0nGS by elotroalex@hcommons.social
2023-03-08T14:05:27Z
0 likes, 0 repeats
@TedUnderwood such believers in thought. It's like a religion with these guys. I'm here for the people who believe in writing. Those people will ask the right question about AI. Instead of saying that the machine doesn't know when something is not the case, they await the moment when the machine can consistently *write* what is not the case.
(DIR) Post #ATcPxtuFi63G5Y60oK by elotroalex@hcommons.social
2023-03-14T21:23:38Z
0 likes, 0 repeats
@TedUnderwood same
(DIR) Post #ATe3dDqLU0PTHhxm3E by elotroalex@hcommons.social
2023-03-15T15:07:29Z
0 likes, 1 repeats
Hat tip to @shannonmattern for sharing a reference to an edited volume on... Box Studies. Shannon, you continue to be the go to person for the epistemologies of that which we were not paying attention to, but maybe we should have. I don't know how you do it, but I'm grateful you are around. You are a gift. Ok, back to reading this fascinating volume.
(DIR) Post #AUVjD61pgFywP8FoFk by elotroalex@hcommons.social
2023-04-10T12:16:24Z
0 likes, 0 repeats
On the plane back from Rome wrote a draft for a grad Fall 2024 course—to be co-taught with Ayesha Ramachandran (if she doesn't run in the opposite direction and avoids me in the hallways from now on once she reads my notes 😃).Here be the course description. A teaser trailer. What say you? Total madness? He's really lost it this time?
(DIR) Post #AUVjEbjfHRIp20TczA by elotroalex@hcommons.social
2023-04-10T13:48:42Z
0 likes, 0 repeats
@CitizenWald best I can do is teach you 😉
(DIR) Post #AUVk4YvkxDpvOFuGWG by elotroalex@hcommons.social
2023-04-10T13:55:45Z
0 likes, 0 repeats
@CitizenWald lol. i just realized that "teach you" could mean be your teacher or teach your materials. I meant the latter.
(DIR) Post #AUVkVFSZSCrYozypIe by elotroalex@hcommons.social
2023-04-10T14:02:54Z
0 likes, 0 repeats
@CitizenWald @bookish been thinking about our colleagues at Rutgers all day today. Godspeed to all. We're rooting for you!
(DIR) Post #AWGnZUGqpM9OevyYU4 by elotroalex@hcommons.social
2023-06-02T01:05:17Z
0 likes, 1 repeats
Prediction:The best generative AI applications will be domain specific. They will combine curated authoritative reference data + LLMs + training by domain experts. Other savvy folk are arguing for this, and prototypes are showing up here and there. What does that mean for historical/cultural scholarship? Bad news. The "curated authoritative reference data" for us means primary and secondary sources. These are mostly in the hands of a few knowledge cartels. It is very likely that they will be the first to drop good AI solutions for us, as long as we do the "training by domain experts." Once more, we will be helping someone else make a buck instead of ourselves. We did it already. We'll do it again. Like suckers.Why do you think these knowledge cartels exist in the first place?
(DIR) Post #AX9Kv1uQO3E1ndpAOG by elotroalex@hcommons.social
2023-06-27T16:13:57Z
0 likes, 0 repeats
My response to Lisa Siraganian "On Accidental and Parasitic Language" on the Critical Inquiry forum is up. Her position statement is clear and inviting. I disagree with her. I do regret the bit of overconfidence and snark in my answer. It comes from a mix of respect + joy. Promise. I'll see myself to the door if y'all think I'm being disrespectful. https://critinq.wordpress.com/2023/06/26/on-accidental-and-parasitic-language/
(DIR) Post #AX9Kv3vCu72W2eCot6 by elotroalex@hcommons.social
2023-06-27T21:22:02Z
0 likes, 0 repeats
@jose_eduardo that is a relief to hear. I don't most of these people. I'll reserve all my snark and overconfidence for @TedUnderwood. He can take it