The AI Apocalypse Is Closer Than You Think?

Comments · 2K

• RE

  @Redarr-i5p · 2 weeks ago

  AI doesn't need to destroy humanity. Humanity is doing a good job doing that to itself.

  628

• BR

  @BRichard312 · 2 weeks ago

  As a Software Engineer turned Data Scientist, I&apos;ve been in the IT game for nearly a quarter century. I&apos;ve worked for several organizations, both private, and public, &nbsp;both as an employee and as a contractor. NOT ONE time did I ever advocate, or develop a sufficiently complex data process model end-to-end WITHOUT some kind of direct oversight.<br><br>When it is being suggested that AI could destroy humanity, what is really being stated, from a technical perspective, is that through reinforcement learning, along with complete agentic automation, WITHOUT oversight, that when this technology has access to the internet, really bad things can happen.<br><br>It all goes back to the first principle of automation. Do NOT automate anything end-to-end that is sufficiently complex WITHOUT some kind of direct oversight. If there is no oversight to agentic AI, really bad things can, and will happen.

  375

• BE

  @beyondrejection1 · 2 weeks ago

  I think someone said it best: “if this model is so smart, why don’t OpenAI and anthropic just use it to make billion dollar businesses themselves”

  262

• WI

  @wilkinru · 2 weeks ago

  In the 20s guys are passionate and still care about stuff. As a 46 yr old...that hits.

  258

• CA

  @calebpearlmusic · 2 weeks ago (edited)

  Your AI company is tens to hundreds of billions of dollars in the hole with no plans on how to make a profit. What do you do?<br> <br>You work with your competition (who is in the same boat) to scare the shit out of the government into tightly regulating the industry to price-out startups and kill competition before it starts because you can&apos;t afford to share the market. Then, you can say you couldn&apos;t develop AGI because of regulation, when you never were able or were going to build it in the first place.

  426

• JO

  @joe-jt5ks · 2 weeks ago

  we got Terminator reenactment before GTA6 ever launch ☕💀

  57

• RO

  @rodtamayoleal · 2 weeks ago

  I&apos;ve read about this being a coordinated effort to prevent the finacial market from collapsing due to the AI bubble bursting very fast, these kind of news: &quot;AI is going to kill us soon&quot;, &quot;AI agents hacked major companies by itself&quot;, etc gives the AI companies an excuse to slow down or even stop development without saying: &quot;Hey guys, we spent trillions and only made millions.... this didn&apos;t went as expected&quot; and bursting the bubble

  76

• RO

  @RoflMaiWaffel · 2 weeks ago

  If he was really blowing the whistle, they&apos;d make him disappear like Suchir Balaji

  100

• SU

  @SuperFx89 · 2 weeks ago

  I&apos;ve heard you can run it in a loop so it generates stuff continuously without being prompted. But yeah, if AI destroys humanity, it&apos;s actually humanity destroying humanity, cause that&apos;s where the training data comes from.

  47

• AB

  @Abdennour_Ab · 2 weeks ago

  Very interesting ! Thanks for sharing !

• MI

  @mikemorris2998 · 2 weeks ago

  AI has already destroyed humanity with all the slop on LinkedIn Asian Dad.

  129

• ST

  @StudyWithTri · 2 weeks ago

  Quality content, thumbs up. I like (agree with) the sentiment that the danger is in the harnesses ie. the tooling and layers enabling agents to interact with the outside