Announcement

Collapse
No announcement yet.

America & the World V: DOGE Business

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    I digress.

    Comment


      In the early days I was more forgiving when it comes to protest votes, it made more sense why people might vote a wild card in protest. But say after the last 13 years of Conservatives was leading into a General Election next year where the only two options were committing to another half decade of Tory insanity or some UKIP/Trump like party or figure - I'd not hesitate to vote for more of the same.

      Left parties might not have done enough whilst in power but the protest vote to vote for something even further right just sends the wrong message and leads to a doubling down. It's more likely that the left leaning party will see the result as an indicator that the public supports less left leaning policies and leadership and just swing harder that way, the same way that left wing voters now dislike how centre Labour has swung.

      I'm not excusing the politicians heavy role in these trends and the poor decisions they make whilst in power, but ignoring the clear evidence elsewhere of how this type of voting pans out and then allowing/enabling it to happen in your own country comes with a level of voter responsibility too. The same as how those who voted Trump in 2020 bear none of the excuses voting in 2016, the ignorance is long gone and armed with the clear evidence/knowledge of what that vote counts toward comes with some of the blame too.

      Voting is one of the few ways the public can voice their discontent with a party but as well meaning as the intent might be to let the prior ruling party know that they're not happy with things might be - proactively voting for something much worse will backfire in spectacular style in all of five minutes. It's more of a shame that more countries don't see protest non-voting, a mass collapse in voting numbers that sends the message that there is a massive opportunity for a third party to emerge and steal votes. The two party system most countries obsess with is broken and this constant right hand push emphasises the dangers of it because it always swings right

      Comment


        Remember when conservatives used to idolise Chilean president Pinochet, and hold him up as a political role model? They’ll do the same with Milei, even as General Ancap destroys Argentina. And it’ll get very very boring.

        Comment


          Does Javier Milei’s dollarisation plan for Argentina make any economic sense? | Javier Milei | The Guardian
          A glance at the potential road Argentina has voted for at an economic level which seems to most likely either involve a much worse level of pain for its public or Milei abandoning most of what he promised.

          Comment


            Far-right party set to win most seats in Dutch elections, exit polls show | Netherlands | The Guardian
            Another country in decline

            Comment


              Jamie Foxx accused of sexual assault in New York bar | Film | The Guardian

              Comment


                It's being reported - by Reuters...



                ... and I recommend people sit down over this...

                ... that the recent kerfuffle over the CEO of OpenAI is because they have, to their own surprise, achieved more than they anticipated and have created an Artificial General Intelligence, i.e. a "true" form of "weak AI". Unlike existing LLMs, this AI can be given a series of maths problems and calculate the answer (ChatGPT can't really do this; if you ask it a maths problem, it'll answer and may be correct, but it does that by memory, not process - like you might know 5x5 is 25; you don't actually calculate it, you just know it).

                This is an enormous step for AI, because it suggests a certain degree of cognitive understanding that is wholly beyond anything ever achieved before. Also, if OpenAI managed to do this, it would be reasonable for them to **** a brick. It's a discovery on a level that might change the course of human history.

                Comment


                  I'll believe they've created an AGI when I see it. Sounds pretty far-fetched to me.

                  This would basically be like creating the first warp drive - AGI is several decades away at least.

                  There's an absolutely massive gulf between that and LLMs - LLMs are basically just huge statistical databases with weightings (and requires masses of compute - hence needing the MS investment). LLMs are not a stepping stone to AGI, they are very different beasts, calling these AI at all is fundamentally wrong.
                  Last edited by MartyG; 23-11-2023, 15:33.

                  Comment


                    Originally posted by MartyG View Post
                    I'll believe they've created an AGI when I see it. Sounds pretty far-fetched to me.

                    This would basically be like creating the first warp drive - AGI is several decades away at least.

                    There's an absolutely massive gulf between that and LLMs - LLMs are basically just huge statistical databases with weightings (and requires masses of compute - hence needing the MS investment). LLMs are not a stepping stone to AGI, they are very different beasts, calling these AI at all is fundamentally wrong.
                    Given vast computing resources, the new model was able to solve certain mathematical problems, the person said on condition of anonymity because the individual was not authorized to speak on behalf of the company. Though only performing math on the level of grade-school students, acing such tests made researchers very optimistic about Q*’s future success, the source said. Reuters could not independently verify the capabilities of Q* claimed by the researchers.
                    so tldr, we have a rumor that they have an AI that can learn basic maths, and its not been verified.

                    Comment


                      The difference is the LLMs are not creating something from first principles - they're just using a vast network of statistics behind a complex algorithm that weights right and wrong, and slowly gets better as it processes more and more data - it has no understanding of what that data actually represents - to the LLM it's just something that matches a pattern it's learned about.

                      An AGI is different as it's able to derive solutions from first principles as a human thinker would - it's not building from collated knowledge as LLMs are.

                      Comment


                        Originally posted by Lebowski View Post
                        so tldr, we have a rumor that they have an AI that can learn basic maths, and its not been verified.
                        Main thing to consider; the unassuming-sounding nature of the "basic maths" belies the nature of this. It's more that the system demonstrates the ability to understand something new, regardless of how simple that is.

                        Comment


                          And that's the part I have an issue with - I massively doubt it's learning anything new. That's not what LLMs do.

                          What this is, is a great way of garnering more investment money though
                          Last edited by MartyG; 23-11-2023, 15:59.

                          Comment


                            Originally posted by MartyG View Post
                            And that's the part I have an issue with - I massively doubt it's learning anything new. That's not what LLMs do.

                            What this is, is a great way of garnering more investment money though
                            It’s definitely not learning as a thinking, conscious mind would and could, because that’s so far out the real of possibilities for a long time.

                            And yep, must be slow on the investment gathering front, so some good old fashioned bull is the order of the day.

                            Comment


                              Elon Musk To Advertisers: “Go F–k Yourself” – Deadline​
                              And yet there are still people who believe he's not a moron with a rich daddy

                              Comment

                              Working...
                              X