rob_c

6 ms·

61.
by rob_c 7mo ago
> It’s sad to see Ars Technica at this level. This was from a journalist _who_is_hired_as_expert_ at knowing of/about tooling that hallucinates (LLM ((AI)) chatbots). Decides to implicitly trust said technology to write a "hit
62.
by rob_c 8mo ago
The article author and the uploader should _BOTH_ be sentient enough to engage brain and not just ignore it because they feel "it's an abstract concept I'd not get in trouble for when not working in the US or EU".
63.
by rob_c 8mo ago
I... There are parts of the world where certain developers don't understand the way the west tends to work with regard to copyright, or not blindly copying anything that is out there. This however is a very, VERY poor situation when yo
64.
by rob_c 8mo ago
Just for reference, teams is not an astounding success it's forced on the workforce by management who want to pay less. It's a classic management square peg into workforce round hole. Yes I understand sometimes something is better
65.
by rob_c 8mo ago
So you just discovered pca in some other form?
66.
by rob_c 8mo ago
And finally we reach the point where you're not shot for explaining if you invest in ownership after everything is over you have something left that has intrinsic value regardless of what you were doing with it. Otherwise, well just li
67.
by rob_c 9mo ago
I think this is akin to x% of the worker ants doing all the work. Once you get to a big enough scale and have to delegate I'm sure every company hits this. I just wish we didn't have to rely on hiring 100 on paper workers for 5 ex
68.
by rob_c 9mo ago
Which when it leads to abuse it's saving face and when it's incompetence it's saving face. For a competent doctor it's used too let a patient know they're doing their job and an acknowledgement of symptoms. Unfortun
69.
by rob_c 9mo ago
This is basically just a rehash of "trained" DNN are a function which is strongly dependent on the initialization parameters. (Easily provable) It would be awesome to have a way of finding them in advance but this is also just a c
70.
by rob_c 9mo ago
1) routing (mis-)config problem - key of remote exploit. This should always be something people double check if they don't understand how it works. 2) hard-coded secrets - this is just against best practice. don't do this _ever_ t
71.
by rob_c 10mo ago
Outside the realm of the testable isn't worth discussing to experimentalists so might as well be a non quantifiable field. Although sociology is perfectly quantifiable and measurable. Even though arguably the underlying relationships b
72.
by rob_c 10mo ago
> a philosopher turned science journalist, Interesting, I read it as the other way round. I wonder which of the many worlds is correct :p The moon example is painful, but I was assuming to be a "if the tree falls in the forest... ya
73.
by rob_c 10mo ago
> That's a year into its availability in France. Anyone who didn't have their first dose by then probably wasn't getting a dose. You are aware of the "incentives" offered by the French govt? Such wonderful option
74.
by rob_c 10mo ago
Yes, a 1% mortality either way would. Yet for some reason we're focused on just one of the possibly results of the decision tree
75.
by rob_c 10mo ago
Explain how? there is a right answer but you'll probably not get it by relying exclusively on the reported data.
76.
by rob_c 10mo ago
NB: most people choosing not to take it in France tend to fall into the medically at risk, stubborn, or, "so far down the rabbit hole that you probably can't trust these people to make sensible life choices" groups. (This alo
77.
by rob_c 10mo ago
If that's not obvious to you pray you're not over of them... But in seriousness it's management failure to build up debt like that. Either self management, middle management or out of touch management. There's a reason
78.
by rob_c 10mo ago
Exactly. If it can't distinguish between a basic repeat after me ordering how is it going to even get a simple output order correct? Let alone pull apart the strategies themselves
79.
by rob_c 10mo ago
Last time I sat down with one of the groups modelling national food availability their model _needed_ a scratch fs capable of dealing with >1M 4kB files per folder. When asked why not to use a db they replied databases don't work we
80.
by rob_c 10mo ago
They're all awful. The ICL model used to "inform" the UK lockdowns had (probably still has) a serious race condition such that when running multi-threaded that meant all of the timelines had errors of +/-1week... (It&#x2
81.
by rob_c 11mo ago
Refuses to learn tool so tool is broken... There is no problem with python for this. If you hate boiler plate job the club, get llms to generate it for you and move on to doing real work (or get involved in improving the language or librari
82.
by rob_c 11mo ago
On the other hand it's literally never been easier. Seriously, don't so much spend the time learning a discipline as learning to learn with any of the free online LLM tools for any subject and if you've actually got the gumpt
83.
by rob_c 11mo ago
> "I’m living a life of unalloyed privilege" How to scream I'm the worst of online discorse
84.
by rob_c 11mo ago
> I worry about the damage caused by these things on distressed people I worry what these people were doing before they "fell under the evil grasp of the AI tool". They obviously aren't interacting with humanity in a norma
85.
by rob_c 11mo ago
the ultimate pebkac...
86.
by rob_c 11mo ago
How to scream I'm behaving badly online...
87.
by rob_c 11mo ago
Yes, but LLM can barely cope with following the ordering of complex software tutorials linearly. Why would you reasonably expect them unprompted to understand time any better enough to trade and turn a profit?
88.
by rob_c 11mo ago
> guessing 'they're poorly suited' is just that, guessing I have a really nice bridge to sell you... This "failure" is just a grab at trying to look "cool" and "innovative" I'd bet. Anyon
89.
by rob_c 11mo ago
At least a coin is faster and more reliable.
90.
by rob_c 11mo ago
Not just can i guarantee the models are bad with numbers, unless it's a highly tuned and modified version they're too slow for this arena. Stick to using attention transformers in better model designs which have much lower latenci

More ›