00:00No, artificial intelligence is not going to annihilate us, but there are other risks.
00:05Dear reader, if the news of the last few weeks has left you confused and with a feeling of uncertainty about
00:12the future of humanity, don't worry.
00:15That was the goal of those who produced them.
00:18It is also a reasonable response to facts that, on paper, seem contradictory.
00:24There's no way to understand how one day Darío Amodé, co-founder of Antrope,
00:30I led a letter warning that artificial intelligence could destroy humans.
00:35And that, a few days later, Amodé himself announces that 25% of the cloud code, the product
00:42star,
00:43It is being written autonomously by artificial intelligence.
00:47Are we heading towards our inevitable extinction?
00:50No, but there are real risks whose causes are human, very human.
00:56We are not going to reproduce here the scandal that arose when a former Antrope employee posted on his account
01:03X
01:03that the most powerful people within the company fear they are building an artificial intelligence
01:09which could even annihilate humanity.
01:12It's worth remembering that Amodé has been saying something similar for years, and that the letter he published was also signed by Elon.
01:21Musk, leader of XAI, and Sam Altan of OpenAI.
01:25If US industry leaders, and this is no small detail, are calling for regulation
01:31And the President of the United States, Donald Trump, responds that the only necessary safeguard is for him to be as
01:39president,
01:40Why are we writing an editorial that calls for reducing panic to its proper proportions?
01:46Well, because we've seen this story before, and the bombast of Amodé and company seeks to influence the public
01:53The world does not focus on what is really happening.
01:58Note how the reports provided by these companies refer to artificial intelligences that act on their own.
02:05and escape the limitations that humans imposed on them.
02:09In other words, the responsibility doesn't lie with them, but with this abstract figure, difficult to understand and almost still
02:17powerless.
02:18The reality is much more mundane.
02:21The Hogan Face incident, where an OpenAI agent managed to bypass security measures and hack a rival company,
02:28This happened precisely because humans had already done something similar.
02:32And that data was in the material they used to train that model.
02:38Cory Doctorow, an expert on the subject and a critic of technology companies, explained it this way and spoke of commonalities.
02:45They didn't create God accidentally.
02:47They created malicious, autonomous software, and then failed to monitor it.
02:53which resulted in that software doing something that was predictable, I'll close the quotes.
02:58In other words, humans failed and then sold a poetic narrative to avoid taking responsibility.
03:06That doesn't mean there are no risks.
03:09Of course there are.
03:11If we continue to build opaque systems that can operate without oversight,
03:16A malicious actor can do a lot of damage.
03:19The solution is not abstract letters, but quick and effective regulation.
03:26Additionally, what is happening is the creation of a market that incentivizes companies to replace humans.
03:32by artificial intelligence.
03:33An automation process that, we were told, would be offset by the new jobs that would be created.
03:41That has not been the case.
03:42Bill Gates said so in the alarm he recently raised.
03:46And I'm opening quotation marks.
03:47Is self-regulation the most dangerous tool ever invented?
03:51No, thanks.
03:52I close quotation marks.
03:53We need to be talking about taxes on the use of artificial intelligence,
03:57of reserves for human labor,
04:00universal basic income,
04:02of strict laws.
04:04That is the crisis we are already facing.
04:07And if you notice how countries like Colombia have no voice or vote in these discussions,
04:14Everything is being decided between Silicon Valley and the authoritarian Chinese regime.
04:20That power imbalance wasn't invented by artificial intelligence.
04:25We humans did it.
04:28If you agree with this editorial,
04:30Send your 500-word pre-editorial to elespectadoropinion.com
04:44Thank you.
04:46Thank you.
04:48Thank you.
04:48Thank you.
04:49Thank you.
04:52Thank you.
04:52Thank you.
Comentarios