Jump to content

Captainant

Certifiably Surly
  • Posts

    17425
  • Joined

  • Days Won

    6

Captainant last won the day on July 9 2024

Captainant had the most liked content!

Reputation

34040 Surly 1%

About Captainant

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Any sufficiently advanced technology is indistinguishable from magic. If you were born in the era of a hydrodynamic drive being the pinnacle of technology, I could see how this sort of advanced mathematical model could seem magical. But Elon is not the wizard or originator of any of the technology that he is fawned over for. I'd bet your FIL spends a bunch of time on Fox and Twitter and has internalized the cult of personality
  2. Nevermind that they all use the same sources for their radar maps, and we have been defunding weather radar systems and deactivating data sharing systems all over the place. Someone said up thread this is complete outlier, and I fear that these data points are going to be less and less the outlier as we lean harder into fucking the environment and climate
  3. Inb4 the sign is changed to Alligator Auschwitz
  4. Yes yes, anyone staking a radical position is just as lame as the mainstream. I'm familiar with the play you're running from Bernie
  5. It literally got trump elected, so why not give it a whirl? Harness the hatred for the extravagantly wealthy, rather than for brown people
  6. Lol you sound like TahoeHorn talking about trump. This is a weird bit to do on July 4, but go off king
  7. Unless you're prosecuting a trumpist lmfao. How the fuck could you type that sentence on July 4, 2025 with a straight face
  8. Concern troll is VERY concerned.
  9. They are dead-ass saying that they're gonna feed alligators "65 million meals" at this facility. ... The US Latino population is around 65 million people.
  10. If Star Trek is any guide, it'll take a nuclear war for people to back off that bugaboo
  11. TBH what you're describing is what the start of my career looked like, only it was crappy brittle spaghetti code that came from TCS. Nobody in house could maintain it and my first big project was refactoring it into a documented and not dogmatically OOP'd codebase - crazy shit like hundreds of lines of class definition and factories just to enable basic operations on JSON objects. Management never wants to invest in good tech and talent, and will only begrudgingly do so after it's demonstrated THAT is what gives a market advantage. On the flip side, cleaning that mess up has been quite a nice career for me so far.
  12. It's fucking crazy how true this is. There's ZERO critical thinking in that class of person, all they see is dollar signs from cutting workers and they don't give a shit about the plummeting quality of product - whether that's customer service interactions, or more brittle and less secure code, or just generally trashing your market reputation. Crazy thing is though - they're gonna be looooong gone before those chickens come home to roost. That's the MBA way, baby! Harvest short term benefits to get their bonuses and then skip town to a new firm before the long-term consequences are made apparent. I think this is true for really basic, reading comprehension stuff. GenAI is gonna absolutely destroy the data entry industry as it currently stands - but that doesn't mean there's nothing else those folks could do. In an information economy, machine learning is the steam or ICE engine. It's what lets us sift through the vast amounts of data. But just because we came up with steam engines doesn't mean that we didn't need rail workers anymore driving stakes. John Henry could have done other jobs workin' on the railroad
  13. This is a fascinating problem space in LLMs! A broadly trained foundational model is gonna be dogshit at mathematical reasoning, as its main corpus of knowledge is mainly general knowledge and writings. So it sounds right that it would have a lay-person's understanding of those subjects. There are techniques called "fine tuning" where you take a base foundational model (like GPT4 or Claude 4.0 or whatever LLM is your fancy) and then do a differential training run on top and introduce in your own corpus of documents. It's SUPER expensive compared to just doing Retrieval Augmented Generation, however it results in a LLM that has a better "feel" for the "music" of the theory than a generalist LLM. All that to say though - they're still pretty dogshit at correctness. To meaningfully improve upon correctness it would take a true Generative Adversarial Network where you're training both a generative model and a "judge" model (academically, it's called a "discriminator" model) that decides if the generated output is dogshit or not. It's can result in very finely tuned and believable generative models, however they have an extremely narrow domain of utility and still take a SHITTON of electricity to produce. And to even start building a GAN architecture, it takes a mountain of data to train the generator and disciminator. Data that doesn't really exist as of right now, and as you correctly point out, is probably going to be generated with AI and result in a Hapsburg AI model
  14. It is my vibe that the higher ed management pipeline is as dog shit and reliant on consulting as every other management pipeline that I've come across. Just there to make a splash and leave before everyone gets wet kind of thing. Am I off base? I feel like everything is boiling down to profit and loss, even education and civic services
×
×
  • Create New...