Roberto

492 posts

Roberto

Roberto

@RobJ02

Econ student, 23. Christian. Philosophy, Cogsci and ML nerd. I also try to keep up with Hebrew Bible and New Testament studies. ES/EN

Peru Entrou em Temmuz 2021
1.5K Seguindo171 Seguidores
Roberto
Roberto@RobJ02·
@xuanalogue "Souls without selves" seems a coherent position approximating some articulations of the buddhist view, there's no essence of what you are in the psychological sense, but there's still a stream of experience and yours is different from everyone else's
English
0
0
1
34
xuan (ɕɥɛn / sh-yen)
xuan (ɕɥɛn / sh-yen)@xuanalogue·
Admittedly the "soul" / "spirit" language is confusing here. Doctrinally there is no "independently-existing soul" in Buddhism, just a continuum of mental states. But that continuum is also understood to be independent of the body (i.e. Buddhism is idealist not physicalist).
English
1
1
12
334
Roberto
Roberto@RobJ02·
@AaronBergman18 @ryanburge @GhostCoase I'd suggest reading @RalphStefanWeir on substance dualism+idealism; from the hard problem he motivates a case against property dualism and for a specific account that gives content to the term. His interviews in Majesty of Reason and Emerson Green are p great
English
0
0
3
107
Aaron Bergman 🔍 ⏸️ (in that order)
Aaron Bergman 🔍 ⏸️ (in that order)@AaronBergman18·
“Soul” is a weirdly vacuous or ambiguous concept under inspection Like do I think there’s something about me above and beyond the physical? Well it’s complicated bc the concept of “physical” also falls apart somewhat under inspection, but to a first approximation: yes, phenomenal consciousness Maybe you wanna call that a “soul” or “spirit,” fine, but I’d prefer to just talk about more specific and substantive claims
English
4
0
23
1.6K
Ryan Burge 📊
Ryan Burge 📊@ryanburge·
I find it fascinating how huge majorities of almost every group agrees that: People have a soul or spirit in addition to their physical bodies. Even 69% of agnostics agree with that. The huge outlier are atheists. Just one-third think that they have a soul.
Ryan Burge 📊 tweet media
English
142
64
766
305.9K
Roberto
Roberto@RobJ02·
@akoustov @petergostev If one expects jaggedness to persist and believes steering will matter for obtaining high-signal work from agents, it seems desirable to have some cheaply evaluable signal of effort
English
0
0
1
23
Roberto
Roberto@RobJ02·
@akoustov Slop exists, and even highly capable models can reinforce ludicrous ideas (see @petergostev's BullshitBench), with little human input and no meaningful expert knowledge steering. It seems worth flagging and ignoring stuff that does not clear even FP-skewed detectors
English
1
0
0
25
Alexander Kustov
Alexander Kustov@akoustov·
"If AI can do your research, your research was never good." Many people said this in response to my earlier piece thinking it was a dunk. It's not. It's an indictment of much of social science. Ten more theses on AI in academia—including what I got wrong and where we go next.
Alexander Kustov tweet media
English
8
15
59
14.3K
Roberto
Roberto@RobJ02·
@rburhum No estoy en desacuerdo con la mayoría de las cosas que afirmas; estoy simplemente señalando que la dicotomía esencialista de términos es otro failure mode en contextos de política. Alguien podría concluir que para responder la segunda conversación debemos ignorar la primera
Español
0
0
0
32
Ragi Yaser Burhum
Ragi Yaser Burhum@rburhum·
Cuando te sientas en una reunión de políticas nacionales o de la ONU: - conversación sobre herramienta: "¿Cómo pagamos para adquirir herramienta específica [x]? ¿Cómo enseñamos como usarla? vs. - conversación sobre infraestructura: "¿Cómo desarrollamos nuestra infraestructura de AI? ¿Cómo la mantenemos y qué tipo de gobernanza necesitamos? Dos conversaciones totalmente distintas.
Español
1
0
1
63
Ragi Yaser Burhum
Ragi Yaser Burhum@rburhum·
ChatGPT es una herramienta bonita, pero el error más grave que cometen los gobiernos al implementar una política de IA es describirla como una herramienta. No lo es. Claro, una lámpara también es una herramienta, pero la electricidad es infraestructura. La IA es infraestructura cognitiva que se está integrando transversalmente en todos los aspectos de un país. EE. UU. y China lo entienden perfectamente. A nivel país, gobernanza, soberanía digital, asimetría entre países, energía, dependencia de proveedores, etc.: esas son las conversaciones que se deben tener cuando se habla de políticas nacionales, no si tienes licencias de Claude o cómo educas a las personas para usar ChatGPT. Lamentablemente, lo común en las discusiones gubernamentales en LATAM sigue siendo ver la IA como una simple herramienta. Terrible error.
Español
10
42
190
10.5K
Roberto
Roberto@RobJ02·
@rburhum No has entendido mi punto, todo lo que dices sobre infraestructura es compatible con pensar en la IA como herramienta en algunos contextos, that's it; no hay una esencia en el término "herramienta" que implique la negación de lo que acabas de escribir
Español
1
0
0
41
Ragi Yaser Burhum
Ragi Yaser Burhum@rburhum·
Hola Roberto. Te explico la importancia: Pensar en la IA como infraestructura en lugar de solo una herramienta cambia la forma en que un país planifica su política económica, su seguridad nacional y su competitividad. Las herramientas se pueden importar; la infraestructura normalmente debe construirse, controlarse y mantenerse dentro del mismo país. La infraestructura permite que surjan ecosistemas completos de herramientas. Si un país solo desarrolla herramientas pero no controla la infraestructura, depende 100% de quienes sí la controlan. Esto va a ser mucha más significativo en el campo de IA. Muchos todavía no se han dado cuenta todavía, pero lo tendrán claro en 10 años. Mas allá del hecho que la infraestructura captura la mayor parte del valor económico, o las consideraciones de seguridad nacional, o el rol que tiene en innovación a largo plazo, estamos hablando de un cambio más radical que la invención de Internet. Tratar esto como un “avance más de tecnología con nuevas herramientas” es una visión muy miope que va a tener consecuencias graves en el futuro. Saludos
Español
1
1
1
177
🗿 
🗿 @sxndrxxg·
@RobJ02 @contracultural No confío mucho en los “detectores de IA”. Lo que detectan en realidad son secuencias de palabras que son probables de suceder una tras otra. Al poner textos antiguos, por ejemplo salta la alerta de IA igualmente.
Español
1
0
0
34
Roberto
Roberto@RobJ02·
@contracultural Sería interesante que puedas preguntarle cómo armó su plan y qué uso dió a la IA; es implausible negar su uso dada la baja tasa de falsos positivos de Pangram, la pregunta es el rol que tuvo en la elaboración pangram.com/history/81fbc0…
Español
0
0
0
64
Roberto
Roberto@RobJ02·
@contracultural Moya, corrí todos los planes de gobierno por Pangram, hay un hilo en mi perfil que quizá te interese. El plan de Nieto tuvo 58%, esos números son consistentes con investigación+iteración con IA, pero también son consistentes con generación total esporádica, ctrl c + ctrl v
Español
2
0
0
291
Consejo de Redacción de Efe
Consejo de Redacción de Efe@cderedaccionefe·
🚨 El periodismo riguroso no puede estar a merced de errores algorítmicos. Desde el Consejo de Redacción de @EFEnoticias denunciamos cómo la IA de Grok ha atacado nuestra credibilidad con información falsa y una rectificación invisible. 🧵 Abrimos hilo.
Español
907
7.5K
13.5K
1.3M
Roberto
Roberto@RobJ02·
@captgouda24 Steep/sustained changes in the prices of commodities, all kinds of labor legislation that increase relative cost of humans, accident(s) at some high risk operation(s), trade changes that turn an agriculture sector from final product to input to other supply chains
English
0
0
1
180
Nicholas Decker
Nicholas Decker@captgouda24·
Hey folks, can you help me brainstorm? I am trying to think of examples of plausibly exogenous shocks to peoples beliefs that their job will be automated, unrelated to technological progress. This could also be on the small-scale -- a plant closure, for instance. Please help!
English
40
5
171
38.4K
Roberto
Roberto@RobJ02·
@TheZvi I think the scheme worried folk have in mind is more like OAI being more willing to appear as frictionless to a regime that operates in the grey zone; resulting in a strategic advantage over a rival company in external guardrails, increased influence in governance and the like
English
1
0
3
525
Zvi Mowshowitz
Zvi Mowshowitz@TheZvi·
Correct. This was never about the money for either OpenAI or Anthropic.
English
11
4
172
17K
Roberto
Roberto@RobJ02·
@CharlotteFang77 @____NanaKong @tszzl I'm mostly worried about the SCR issue, I don't think any public information justifies that measure. I understand your criticisms, but ultimately disagree
English
0
0
9
489
♡ Charlotte Fang 🪲 Crown Prince ❀ LOVE HEALS 💞
♡ Charlotte Fang 🪲 Crown Prince ❀ LOVE HEALS 💞@CharlotteFang77·
If you're paying any close attention and can swallow your biases, the OAI announcement made more evident Anthropic's PR has been misleading about what terms were actually under contention, which would justify a retraction of early support unrelated to professional affiliation. Regardless, it remains poor form for the reasons stated above, you're poisoning the commons.
English
6
0
3
618
Roberto
Roberto@RobJ02·
@tszzl’s tweets, now deleted, seemingly minutes before learning of OpenAI’s deal with the DoD. See specifically the second
Roberto tweet mediaRoberto tweet media
English
26
133
2.1K
281.4K
Roberto
Roberto@RobJ02·
@CharlotteFang77 @____NanaKong @tszzl I agree that this effect is possible; and favor taking them into account, I just think considerations in this case warranted sharing. It's an apparent dissonance between beliefs of someone working in a lab and the actions of its leadership that warrants attention given the stakes
English
1
0
5
636
Roberto
Roberto@RobJ02·
@____NanaKong @tszzl Those of us less bullish on OAI's motivations will of course make different judgments and think it important to document attitudes expressed publicly, and not just reproduce the company's narrative. This shouldn't turn into vicious behavior tho, I'm hiding replies in that spirit
English
1
0
20
1.5K
Roberto
Roberto@RobJ02·
@____NanaKong @tszzl I agree with what you wrote about him being "professional" here; given his mental model of the situation it makes sense for him to choose restraint, delete the tweets, be cautious and even optimistic. I think he's been consistent on the SCR issue, for example
English
1
0
29
22.3K
Roberto
Roberto@RobJ02·
@tszzl @UnderSecretaryF "We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements"; how is that consistent with OpenAI leadership; communicating, tacitly, that Anthropic should've just accepted DoW's terms?
English
0
0
2
124
Roberto
Roberto@RobJ02·
@tszzl the second part of that first sentence is important though, "..which in our opinion we think everyone should be willing to accept". Specially in light of @UnderSecretaryF's tweet you just shared that specifies that "this...is a compromise that Anthropic was offered, and rejected"
English
1
0
6
939
roon
roon@tszzl·
most important part of this >We are asking the DoW to offer these same terms to all AI companies [...] We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements.
Sam Altman@sama

Tonight, we reached an agreement with the Department of War to deploy our models in their classified network. In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome. AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement. We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only. We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements. We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place.

English
169
19
842
172.8K

Descobrir

@xuanalogue @AaronBergman18 @ryanburge @GhostCoase @RalphStefanWeir @akoustov @petergostev @rburhum