• 0 Posts
  • 8 Comments
Joined 3 years ago
cake
Cake day: July 16th, 2023

help-circle


  • I wish that was the case for me. Unfortunately, even when I get internal emails about AI in science events at best it’s unclear. But often it’s generative AI. At this point I’m kinda just forced to assume it means genAI unless stated otherwise.

    Though I guess part of that is that most research groups will specify the kind of machine learning they use. So if it’s left as AI then it’s usually generative.

    Outside of that though it is true that (as seen in OP) most news articles will just label any model as AI, diluting the term entirely. And it’s probably a neural network if it’s from a reputable research group (such as in this case: a University of Cambridge group running actual human trials).


  • GB News… Yuck. Though good job, looks like the BBC has almost the same article.

    Either way, looking at the actual published paper, they used DIOSynVax’s tech. No mention of LLMs or AI anywhere. All I can find is “computational biology” and “self-improving algorithms”.

    In other words: this is using the current genAI boom to hype up/get clicks on something almost certainly designed using well-established machine learning algorithms. Oniony headline I guess, but otherwise it’s just bad journalism in my opinion.

    God I fucking hate the way the term AI is used for everything now. Usually it means an LLM or some other form of generative model in laymen’s circles, but people so happily mix it up with other tools.



  • I’m seeing this post a bit late, but I feel like I have to weigh in slightly, though it’s not my research area.

    Note that my information extends more to academic studying, don’t know if it’s quite as true for learning more physical skills.

    The main concept for learning is deeper learning. Which basically just means actually using your brain to think about the material. Things like connecting it to other ideas, pondering different implications, that sort of thing.

    The reason flashcards work is because you think about what questions you could ask about the material. The reason you write by hand vs type is because it’s slower and you have to think about what’s more important or how you’d summarise the information.

    I believe reading aloud typically works because it forces you to be slower and more deliberate, giving you time to actually process what you’re reading.

    That said what you’ve written is helpful and mostly correct, I’m just not so certain about the framing. It could mislead some people into just rewriting notes while reading them out, for example, which is inefficient and not very helpful for learning.

    A very easy-to-read source with practical tips:

    • Optimizing Learning in College by Putnam et al. (2016) (Look it up on Google scholar for a free pdf)

    Also as a final tip, my favourite exam prep technique: do a past paper without having looked at any notes or done any prep. Answer as much as you can just thinking about what you remember. Then go through with notes. It primes your brain for processing and storing the information.



  • Wimopy@feddit.uktoMemes@sopuli.xyz•Unrealistic
    link
    fedilink
    arrow-up
    0
    ·
    8 months ago

    Just to nerd out a bit: I think this is justified. He has nanomachines, son. His eyesight is probably perfect, but he adopts a different persona most of the time. He’s a puppeteer and a politician most of the time.

    This is when he gets his actual boss theme lyrics too and he finally treats you as an actual fight. No more metal gear, no more speeches. You get the treatment of an equal rival.

    Or: Rising Revengeance might like symbolism, but it prefers the subtlety of a giant boulder and memes, and him taking off his glasses just fits the rule of cool.