haldujai

52 ms·

571.
by haldujai 3y ago
I just downloaded it, looks pretty useful especially for news sections I don’t follow particularly closely. Will check this out over a few days. Out of curiosity have you looked at deep you can go while still getting quality results (i.e. b
572.
by haldujai 3y ago
I don’t think it’s a given that finetuning address pretraining bias, most LLMs you would have access to are trained on very similar and biased corpora, see any Anthropic or OpenAI alignment paper. It’s conceptually possible for extensive fi
573.
by haldujai 3y ago
A black-box LLM which is probably inherently biased from its training data ascribing racist labels to individual humans and ranking them sounds like a horrendous idea you’d read about in an Orwell novel. Do you have any evidence suggesting
574.
by haldujai 3y ago
Summarized bullshit is still bullshit akin to a polished turd. Given that the choice of which articles to write is incredibly biased to begin with this approach does not seem effective. What could theoretically work is an “AI news agency” t
575.
by haldujai 3y ago
Are you curating the list of publishers somehow? I would imagine the AP/AFP newswire repost circuit/echo chamber would result in overestimating importance of a lot of crapola and underestimate the importance of investigative piece
576.
by haldujai 3y ago
Is that even possible? Take the specific language apart the choice of which stories to report on is already/can be highly biased. Not sure a sanitized Fox News or NYT is “bias free” and is probably more dangerous than approaching the o
577.
by haldujai 3y ago
They actually use a 5-dimension political spectrum. Left-right is not specific to any political party, dates back 200+ years and is extensively used in academic and nonacademic literature. What other system would you recommend for evaluatin
578.
by haldujai 3y ago
Breakdown of functional integration =/= different parts of the brain can independently conduct behavior and gain knowledge. Still not sure what the contradiction is, is it regret now? Cause that isn’t contradictory. It’s pretty simple,
579.
by haldujai 3y ago
I assume this is just a PR/IR-driven project to stay the "Google is Dead" headlines hence the budget, especially considering an oversized chunk was spent on the scaling law, doesn't seem they were serious about building
580.
by haldujai 3y ago
I wonder how special these architectures are compared to what's published. The "secret sauce" may just be getting 2 pages (~200) worth of engineers collaborating and either rolling out your own cloud service or spending $$$ a
581.
by haldujai 3y ago
Depends on your goal, if it's to overtake OpenAI as having the best model overall it makes sense to optimize for training loss alone (assuming a fixed upfront compute budget). Optimizing for inference to achieve the same loss would req
582.
by haldujai 3y ago
I'm arguing against the strength of your statements based on methodologically unsound experiments that do not "very very clearly indicate" anything beyond pose a few questions for which there are several different hypothetica
583.
by haldujai 3y ago
You're ignoring several confounders and conflating several loosely related things that I'm not even sure what the point you're making is anymore. To begin with, the split-brain experiments don't provide clear or strong e
584.
by haldujai 3y ago
Says who? There is actual evidence to support that our brain doesn't "know" what it is doing on a subconscious level? As far as I'm aware it's more that conscious humans don't understand how our brain works. I
585.
by haldujai 3y ago
I'm reading this as our explanations for our own behavior as in why am I typing on this keyboard right now, in which case it's not evident at all. The existence of cognitive dissonance suggested in your citation is in no way analo
586.
by haldujai 3y ago
Agree I'm not a neuro subspecialist but I've listened to some talks at conferences out of interest and I don't think anyone still believes in this anymore. Anecdotally the few fMRI's I reported as a trainee didn't s
587.
by haldujai 3y ago
I'm not understanding the connection between your paragraphs here even after reading the first article. Even if you accept classic theory (e.g. hemispheric localization and the homunculus) which most experts don't all this suggest
588.
by haldujai 3y ago
This graph only shows consumer banking failures which captures a tiny fraction of the financial institution failures in 2008. Additionally, this dataset by definition doesn’t consider the several financial institutions (including much large
589.
by haldujai 3y ago
If we're comparing the Great Recession which was not caused by or relatively impacted consumer banks it seems arbitrary to limit the dataset to them.
590.
by haldujai 3y ago
This is not even remotely close to 2008… This figure conveniently excludes the largest 6 failures of 2008, TARP, and major international banks and insurers simultaneously affected. As a reference the prospect of AIG failing without a bailou
591.
by haldujai 3y ago
If two regional banks failing with little realized loss is massive and unprecedented how would you describe 2007-2008 when the entire world economy was brought to its knees with napkin math ~1.2T of bailouts/failures in the US let alon
592.
by haldujai 3y ago
The difference is this is missing the biggest failures of 2008 (Lehman, Bear, Merrill) as well as Fannie, Freddie, AIG and TARP.
593.
by haldujai 3y ago
Nominally yes, but not when you consider borrowing cost and effects on deficit. The CBO has TARP costing 31 billion. [0] FDIC is paid by the banks but it seems probable it will be passed down to customers in banking fees at the end of the d
594.
by haldujai 3y ago
Agree, this representation also makes WaMu’s failure look like the worst in recent history but it felt like one of the smaller problems at the time with what was going on with the investment banks, Fannie/Freddie and AIG.
595.
by haldujai 3y ago
At the very least that’s missing Fannie, Freddie, Bear, Merrill, Lehman, TARP and arguably AIG for another 1.2T+, granted a lot of this was eventually repaid as the FDIC will be as well.
596.
by haldujai 3y ago
You are completely correct with the minor edit that this study would never be shutdown as there is a 0% chance it ever gets approved. Beyond how strict IRB/REB boards are in general, there are also additional federal requirements that
597.
by haldujai 3y ago
What studies? The link you shared is a click-baity ("Science stopped believing in X and you should too") article by an individual clinical psychologist referencing (and overstating) the findings of a proposed etiological model and
598.
by haldujai 3y ago
It's unequivocal that the state has the right and obligation to intervene in cases of child neglect/abuse (which refusal or inability to provide necessary medical care is an example of) and is easily evidenced by the existence of
599.
by haldujai 3y ago
I'm unclear whether you're arguing that obscene/pornographic material should be accessible/distributed to minors as the distinction here is that this activity is in fact a federal crime in the US (and most of the world).
600.
by haldujai 3y ago
You're a bit overly focused on semantics here and misinterpreting the rationale behind it. Behavioral addictions are acknowledged in DSM-5 although this does not extend to compulsive sexual behavior at this time due to insufficient (ye

More ›