Yosha Iglesias
@yoshaiglesias.bsky.social
2025 French Chess Champion, Woman International Master, Fide Master, chess coach, chess composer, YouTuber 🏳️⚧️
[email protected]
[email protected]
or "maybe we should listen to all these scientists warning us about the threat of AI and at least have some reasonable doubts" you get insulted and threatened.
I see the cons of staying here, and I don't think the pros outweigh them.
I see the cons of staying here, and I don't think the pros outweigh them.
September 10, 2026 at 7:39 AM
or "maybe we should listen to all these scientists warning us about the threat of AI and at least have some reasonable doubts" you get insulted and threatened.
I see the cons of staying here, and I don't think the pros outweigh them.
I see the cons of staying here, and I don't think the pros outweigh them.
I really don't care about @fireantprincess.bsky.social whom I didn't know yesterday. But it's so symptomatic of my experience on Bluesky. You can say random stuff or stuff the vast majority here agrees on.
But if you dare say something like "more violence is not a good way for having less violence"
But if you dare say something like "more violence is not a good way for having less violence"
September 10, 2026 at 7:39 AM
I really don't care about @fireantprincess.bsky.social whom I didn't know yesterday. But it's so symptomatic of my experience on Bluesky. You can say random stuff or stuff the vast majority here agrees on.
But if you dare say something like "more violence is not a good way for having less violence"
But if you dare say something like "more violence is not a good way for having less violence"
Nevertheless, I tried to have a normal conversation. Maybe confronting points of view could be a way to learn or teach a few things?
When for some reason Charlotte brought up one of my chess posts I had some hope.
But no, the point here is not to debate, but to insult people who disagree with you
When for some reason Charlotte brought up one of my chess posts I had some hope.
But no, the point here is not to debate, but to insult people who disagree with you
September 10, 2026 at 7:39 AM
Nevertheless, I tried to have a normal conversation. Maybe confronting points of view could be a way to learn or teach a few things?
When for some reason Charlotte brought up one of my chess posts I had some hope.
But no, the point here is not to debate, but to insult people who disagree with you
When for some reason Charlotte brought up one of my chess posts I had some hope.
But no, the point here is not to debate, but to insult people who disagree with you
That was my "crime".
We all react differently to violence. I can't stand it.
I can understand violence when it comes from fascists because fascism is by nature violent, but I can't understand it when it comes from people supposedly on my side in the global war against fascism.
We all react differently to violence. I can't stand it.
I can understand violence when it comes from fascists because fascism is by nature violent, but I can't understand it when it comes from people supposedly on my side in the global war against fascism.
September 10, 2026 at 7:39 AM
That was my "crime".
We all react differently to violence. I can't stand it.
I can understand violence when it comes from fascists because fascism is by nature violent, but I can't understand it when it comes from people supposedly on my side in the global war against fascism.
We all react differently to violence. I can't stand it.
I can understand violence when it comes from fascists because fascism is by nature violent, but I can't understand it when it comes from people supposedly on my side in the global war against fascism.
But it's not. Some time ago I was threatened and insulted for ironically expressing the regret that so many on the left no longer believe in pacifism.
Yesterday @fireantprincess.bsky.social threatened and insulted me for saying we should listen to scientists telling us AI might become a real threat
Yesterday @fireantprincess.bsky.social threatened and insulted me for saying we should listen to scientists telling us AI might become a real threat
September 10, 2026 at 7:39 AM
But it's not. Some time ago I was threatened and insulted for ironically expressing the regret that so many on the left no longer believe in pacifism.
Yesterday @fireantprincess.bsky.social threatened and insulted me for saying we should listen to scientists telling us AI might become a real threat
Yesterday @fireantprincess.bsky.social threatened and insulted me for saying we should listen to scientists telling us AI might become a real threat
Despitethere being a few prople I like, there are basically no chess players, and no impact to have on the chess world.
Chess is not all and being in a queer friendly and leftist safe place could be nice for a change, I thought.
Chess is not all and being in a queer friendly and leftist safe place could be nice for a change, I thought.
September 10, 2026 at 7:39 AM
Despitethere being a few prople I like, there are basically no chess players, and no impact to have on the chess world.
Chess is not all and being in a queer friendly and leftist safe place could be nice for a change, I thought.
Chess is not all and being in a queer friendly and leftist safe place could be nice for a change, I thought.
I've been assaulted a few times in my life and even though the threats on Twitter never felt like they could materialize, they sometimes had a real impact on my mental health.
But it was easy to think that fascism is based on violence and use that as a daily reminder of what I'm fighting against.
But it was easy to think that fascism is based on violence and use that as a daily reminder of what I'm fighting against.
September 10, 2026 at 7:39 AM
I've been assaulted a few times in my life and even though the threats on Twitter never felt like they could materialize, they sometimes had a real impact on my mental health.
But it was easy to think that fascism is based on violence and use that as a daily reminder of what I'm fighting against.
But it was easy to think that fascism is based on violence and use that as a daily reminder of what I'm fighting against.
Not to mention the computing power to create the tablebase...
So unless we have many technological revolutions, it will never happen.
So yes, AGI and SGI will happen much sooner.
So unless we have many technological revolutions, it will never happen.
So yes, AGI and SGI will happen much sooner.
September 9, 2026 at 4:42 PM
Not to mention the computing power to create the tablebase...
So unless we have many technological revolutions, it will never happen.
So yes, AGI and SGI will happen much sooner.
So unless we have many technological revolutions, it will never happen.
So yes, AGI and SGI will happen much sooner.
There are about 10^45 legal chess positions so we need 10^47 bytes
On earth we have about 10^22 bytes
So we need about 10^25 more storage capacity.
That means we need to somehow put and use 10 times our current storage capacity on every star system in the whole universe (~10^24 stars)
On earth we have about 10^22 bytes
So we need about 10^25 more storage capacity.
That means we need to somehow put and use 10 times our current storage capacity on every star system in the whole universe (~10^24 stars)
September 9, 2026 at 4:42 PM
There are about 10^45 legal chess positions so we need 10^47 bytes
On earth we have about 10^22 bytes
So we need about 10^25 more storage capacity.
That means we need to somehow put and use 10 times our current storage capacity on every star system in the whole universe (~10^24 stars)
On earth we have about 10^22 bytes
So we need about 10^25 more storage capacity.
That means we need to somehow put and use 10 times our current storage capacity on every star system in the whole universe (~10^24 stars)
It's very easy to understand, just give it at try:
To fully solve chess (32-pieces tablebase) you need to map every possible chess position to a list of its legal moves and link each legal move to its evaluation.
For each position, you need ~100 bytes to map the moves and their evals
To fully solve chess (32-pieces tablebase) you need to map every possible chess position to a list of its legal moves and link each legal move to its evaluation.
For each position, you need ~100 bytes to map the moves and their evals
September 9, 2026 at 4:42 PM
It's very easy to understand, just give it at try:
To fully solve chess (32-pieces tablebase) you need to map every possible chess position to a list of its legal moves and link each legal move to its evaluation.
For each position, you need ~100 bytes to map the moves and their evals
To fully solve chess (32-pieces tablebase) you need to map every possible chess position to a list of its legal moves and link each legal move to its evaluation.
For each position, you need ~100 bytes to map the moves and their evals
Wow, how self-confident you must be to tell a literal chess master she's "out of touch" about a chess matter
Girl, I promise you there is no way (as in physically impossible as we understand computers) we fully solve chess in our lifetime (and probably ever)
Yes, AGI is many oom easier to achieve.
Girl, I promise you there is no way (as in physically impossible as we understand computers) we fully solve chess in our lifetime (and probably ever)
Yes, AGI is many oom easier to achieve.
September 9, 2026 at 12:52 PM
Wow, how self-confident you must be to tell a literal chess master she's "out of touch" about a chess matter
Girl, I promise you there is no way (as in physically impossible as we understand computers) we fully solve chess in our lifetime (and probably ever)
Yes, AGI is many oom easier to achieve.
Girl, I promise you there is no way (as in physically impossible as we understand computers) we fully solve chess in our lifetime (and probably ever)
Yes, AGI is many oom easier to achieve.
No, this site is one of the basis of my stance. Now if you read this source and think "oh there is 0% SGI will happen in our lifetime", then I understand how you can proudly call me dumb while quoting a source that shows that your previous certainty was totally unfounded.
September 9, 2026 at 11:57 AM
No, this site is one of the basis of my stance. Now if you read this source and think "oh there is 0% SGI will happen in our lifetime", then I understand how you can proudly call me dumb while quoting a source that shows that your previous certainty was totally unfounded.
September 9, 2026 at 11:40 AM
Sentience is a totally different problem than SGI! I personnaly think no program will ever have a subjective experience but that's not the point. SGI will be reached in our lifetimes if we don't prevent it and we won't be able to control it, sentient or not.
You suggestion might be the best way.
You suggestion might be the best way.
September 9, 2026 at 11:36 AM
Sentience is a totally different problem than SGI! I personnaly think no program will ever have a subjective experience but that's not the point. SGI will be reached in our lifetimes if we don't prevent it and we won't be able to control it, sentient or not.
You suggestion might be the best way.
You suggestion might be the best way.
So you can't express a rational, evidence-based, scientific doubt? It's 0% everything is fine until it's 100% oh shit we are already dead?
And wtf cares if SGI is reached with LLMs only, LLMs and strong architecture, world models, or smth else?
The consensus is it will be reached in our lifetime.
And wtf cares if SGI is reached with LLMs only, LLMs and strong architecture, world models, or smth else?
The consensus is it will be reached in our lifetime.
September 9, 2026 at 11:29 AM
So you can't express a rational, evidence-based, scientific doubt? It's 0% everything is fine until it's 100% oh shit we are already dead?
And wtf cares if SGI is reached with LLMs only, LLMs and strong architecture, world models, or smth else?
The consensus is it will be reached in our lifetime.
And wtf cares if SGI is reached with LLMs only, LLMs and strong architecture, world models, or smth else?
The consensus is it will be reached in our lifetime.
So predictable... the right has climate deniers, we have AI deniers.
September 9, 2026 at 10:59 AM
So predictable... the right has climate deniers, we have AI deniers.
(Si tu parlais de la confiance au moment du post de Terence Tao, soit)
September 9, 2026 at 7:36 AM
(Si tu parlais de la confiance au moment du post de Terence Tao, soit)
Je n'ai jamais formellement étudié l'épistémo, mais aujourd'hui n'est-on pas à un degré de confiance qui serait plus similaire à une PJ HC+ vs. C+ ?
(Sauf qu'ici, ironiquement, à l'inverse des PJ on considère que la preuve par ordi seule est douteuse et on demande que des humains la comprennent)
(Sauf qu'ici, ironiquement, à l'inverse des PJ on considère que la preuve par ordi seule est douteuse et on demande que des humains la comprennent)
September 9, 2026 at 7:36 AM
Je n'ai jamais formellement étudié l'épistémo, mais aujourd'hui n'est-on pas à un degré de confiance qui serait plus similaire à une PJ HC+ vs. C+ ?
(Sauf qu'ici, ironiquement, à l'inverse des PJ on considère que la preuve par ordi seule est douteuse et on demande que des humains la comprennent)
(Sauf qu'ici, ironiquement, à l'inverse des PJ on considère que la preuve par ordi seule est douteuse et on demande que des humains la comprennent)
Mais les annonces de ce genre (pour des problèmes moins glorieux et impressionnants) ont jusqu'à présent toujours été vérifiées, non ?
Quel intérêt marketing à faire une fausse annonce ?
Qu'il faille attendre les confirmations par les pairs, je suis d'accord, mais je les vois pas mentir ou bluffer.
Quel intérêt marketing à faire une fausse annonce ?
Qu'il faille attendre les confirmations par les pairs, je suis d'accord, mais je les vois pas mentir ou bluffer.
September 8, 2026 at 10:08 PM
Mais les annonces de ce genre (pour des problèmes moins glorieux et impressionnants) ont jusqu'à présent toujours été vérifiées, non ?
Quel intérêt marketing à faire une fausse annonce ?
Qu'il faille attendre les confirmations par les pairs, je suis d'accord, mais je les vois pas mentir ou bluffer.
Quel intérêt marketing à faire une fausse annonce ?
Qu'il faille attendre les confirmations par les pairs, je suis d'accord, mais je les vois pas mentir ou bluffer.
Du coup tu penses quoi des annonces de ce soir ? Ils se trompent ? Ils mentent ?
(J'ai cru comprendre que dans tous les cas les vérifications officielles peuvent prendre des mois mais ils ont l'air autant sûrs d'eux que tu as l'as sûr de toi)
www.france24.com/en/live-news...
(J'ai cru comprendre que dans tous les cas les vérifications officielles peuvent prendre des mois mais ils ont l'air autant sûrs d'eux que tu as l'as sûr de toi)
www.france24.com/en/live-news...
OpenAI says AI solved one of math's hardest problems in days
OpenAI said Tuesday that one of its unreleased artificial intelligence models had solved a mathematical puzzle that has eluded mathematicians for generations, in a claim which also drew a fight with…
www.france24.com
September 8, 2026 at 9:04 PM
Du coup tu penses quoi des annonces de ce soir ? Ils se trompent ? Ils mentent ?
(J'ai cru comprendre que dans tous les cas les vérifications officielles peuvent prendre des mois mais ils ont l'air autant sûrs d'eux que tu as l'as sûr de toi)
www.france24.com/en/live-news...
(J'ai cru comprendre que dans tous les cas les vérifications officielles peuvent prendre des mois mais ils ont l'air autant sûrs d'eux que tu as l'as sûr de toi)
www.france24.com/en/live-news...
Et transphobe* ( je me permets de préciser, même si établir une liste exhaustive des discriminations en un post aurait été un exploit)
September 8, 2026 at 8:01 PM
Et transphobe* ( je me permets de préciser, même si établir une liste exhaustive des discriminations en un post aurait été un exploit)