[–] 18 points 4 days ago (2 children)

I genuinely wonder what these executives have in their mind.

What do they think? That engineers are good enough for being hired but not good enough to decide which tools to use?

If I hire a plumber and I tell them what tool to use, the plumber would either comply and do a bad job or will tell me to get lost.

Why do they think this is different?

Do some engineers like LLM? Sure! And let it be Do some engineers don’t like LLM? Yes… and let it be.

If these tools are so miraculous the engineers that use them will provide such good results that will naturally prevail.

Why the executives need to thumb the scale?

  • source
  • [–] 4 points 4 days ago

    Definitely buy for your own domain.

    Sadly it’s the only way of having “portability” for mails

    Tuta? Sure Infonaniak? Why not? The next thing? We will see Self hosting? Maybe you’ll have enough time to do that one day

    The only way to have control of your digital destiny is portability

  • source
  • parent
  • context
  • [–] 72 points 2 weeks ago (1 child)

    Totally unrelated: toddlers test adults boundaries. Once the adult give up on a boundary, that is a given and they would put a tantrum if they don’t keep in getting their way.

    What it’s worse is that they will move to the next boundary and repeat the same.

    In general I would say for toddler is better to set clear and firm boundaries so that a lot of tantrum and screaming are saved.

    But what was the article about, again?

  • source
  • [–] 6 points 3 weeks ago (1 child)

    Spies will find other ways to enter.

    This kind of things impact only everyday people. This is kinda the point of this type of measures: the idea is to pressure people to rebel to their governments. But I think Russians are not exactly optimistic this can be done (despite having their fair share of protests that end in Siberians prisons)

  • source
  • parent
  • context
  •  

    It would be interesting to read the post mortem (if they will release). In particular in which phase of the training it escaped the sandbox, in which way it used the information. Training is different from inference, I doubt there is a context there and where would the model store these information? In the weights?! That makes little sense to me

     

    As opposed to the northern/southern hemisphere that has:

    • a clear reference point (equators)
    • traditionally we don’t model a rotation north/south

    Western/eastern hemisphere

    • doesn’t have a clear reference (Greenwich?)
    • traditionally we model rotation on the north/south axis so western/eastern would depend from the time
    view more: next ›