2025-02-21 00:02:16 <veltas      > I mean you look at the roman empire, they had a few good emperors and a lot of bad emperors.  But a lot of those bad emperors were never really in charge anyway.
2025-02-21 00:02:20 <veltas      > Got to look out for those bad emperors that *are* in charge
2025-02-21 00:03:43 <KipIngram   > How do you know not to do something again if you don't know how bad it went the last time it was tried?
2025-02-21 00:04:55 <thrig       > https://thrig.me/tmp/history.jpg
2025-02-21 00:05:25 <veltas      > True that
2025-02-21 00:05:38 <KipIngram   > Oh man.  Yeah.
2025-02-21 00:05:54 <KipIngram   > At least the ones that don't know get to be excited on the way in.
2025-02-21 00:20:57 <veltas      > But how close are we to AGI?  Does anyone here know a lot about AI research?
2025-02-21 00:26:42 <thrig       > someone pinned it to 100 billion in profits, which seems much in the line of "AI as a term invented to get money"
2025-02-21 00:40:20 <nmz-        > AGI requires more than a word pattern guesser
2025-02-21 00:41:00 <nmz-        > when you think, your prefrontal cortex imagines, it plays games, it simulates several realities and then tells you what's going to happen, you then translate that into words
2025-02-21 00:41:28 <nmz-        > LLMs are text based, they don't know what a picture is, they don't know what graphics are
2025-02-21 03:07:14 <KipIngram   > Also, the way you know how to simulate those realities is by having operated in the world - moved your body around, interacted with things, etc. - that's how you know how these simulations are likely to go.  So it may be that real intelligence requires that as well.
2025-02-21 04:03:58 <nmz-        > that's a physics thing then
2025-02-21 04:04:36 <nmz-        > which is a part of the simulated game thing
2025-02-21 04:45:11 <Russtopia   > Hello! Saw the posting on lemmy about the gforth bot!
2025-02-21 04:45:28 <Russtopia   > Now... to learn some gforth :p
2025-02-21 04:45:43 <Russtopia   > !gforth
2025-02-21 04:52:22 <Russtopia   > Uhm... !gforth .s
2025-02-21 05:45:53 <nmz-        > and he gone
2025-02-21 05:45:58 <nmz-        > shame, shame
2025-02-21 06:20:51 <xentrac     > !gforth .s
2025-02-21 06:20:52 <gforth_eval_bot> <0>
2025-02-21 06:21:41 <xentrac     > nmz-: probably nobody knows what AGI requires because probably nobody has built one yet
2025-02-21 06:22:28 <xentrac     > airplanes don't have to flap their wings
2025-02-21 08:24:30 <nmz-        > xentrac: yeah... but at least I can tell via how I think
2025-02-21 09:33:44 <veltas      > I'd expect experts to have predictions for how much longer it will take, with a lot of variability
2025-02-21 09:34:05 <veltas      > Hard to find experts and not just hype monsters, though
2025-02-21 09:45:01 <veltas      > I'm going to have to find a project to do in konilo
2025-02-21 09:45:08 <veltas      > Probably best way to learn it
2025-02-21 12:19:27 <xentrac     > nmz-: if introspection were sufficient to understand the mind, we'd have had AGI about 60 years ago
2025-02-21 12:20:16 <xentrac     > also much less debate about philosophy and psychology
2025-02-21 19:13:29 <Guest47     > !gforth 1 .
2025-02-21 19:13:30 <gforth_eval_bot> 1
2025-02-21 19:37:14 <deadmarshal_> !gforth 1 2 swap
2025-02-21 19:37:16 <deadmarshal_> !gforth 1 2 swap .s
2025-02-21 19:37:17 <gforth_eval_bot> <2> 2 1
2025-02-21 19:43:57 <crc         > !gforth version-string type
2025-02-21 19:43:58 <gforth_eval_bot> 0.7.3
2025-02-21 20:05:12 <pgimeno     > !gforth : testelse if 1 else 2 else 3 else 4 then ;   1 testelse . . 0 testelse . .
2025-02-21 20:05:12 <gforth_eval_bot> 3 1 4 2
2025-02-21 20:05:37 <pgimeno     > !gforth : testelse if 1 else 2 else 3 else 4 then ;   1 testelse . . 124 emit 0 testelse . .
2025-02-21 20:05:38 <gforth_eval_bot> 3 1 |4 2
2025-02-21 20:08:31 <xentrac     > (this behavior is required by the Forth standard, which is hosted on the same web server as the GForth documentation because of the large overlap in contributors)
2025-02-21 20:19:36 <pgimeno     > yes, it was discussed some weeks ago I believe
2025-02-21 20:20:30 <pgimeno     > though the extra info about the contributors overlay is new :)
2025-02-21 20:20:37 <pgimeno     > to me at least
2025-02-21 20:20:56 <pgimeno     > overlap*
2025-02-21 20:23:05 <xentrac     > I noticed this when the server went down for a few days recently
2025-02-21 21:51:10 <Russtopia   > !gforth 1 2 + .
2025-02-21 21:51:11 <gforth_eval_bot> 3
2025-02-21 21:55:20 <lispmacs[work]> !gforth . . .
2025-02-21 22:00:51 <Russtopia   > !gforth words
2025-02-21 22:00:54 <gforth_eval_bot> https://0x0.st/8b2g.txt
2025-02-21 22:49:47 <veltas      > The real standard https://www.iso.org/standard/26479.html