Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 17 submissions in the queue.
posted by Fnord666 on Sunday January 26, @04:38AM   Printer-friendly
from the AI-overlords dept.

https://arstechnica.com/ai/2025/01/anthropic-chief-says-ai-could-surpass-almost-all-humans-at-almost-everything-shortly-after-2027/

On Tuesday, Anthropic CEO Dario Amodei predicted that AI models may surpass human capabilities "in almost everything" within two to three years, according to a Wall Street Journal interview at the World Economic Forum in Davos, Switzerland.
[...]
Amodei co-founded Anthropic in 2021 with his sister Daniela Amodei and five other former OpenAI employees. Not long after, Anthropic emerged as a strong technological competitor to OpenAI's AI products (such as GPT-4 and ChatGPT). Most recently, its Claude 3.5 Sonnet model has remained highly regarded among some AI users and highly ranked among AI benchmarks.
[...]
Even with his dramatic predictions, Amodei distanced himself from a term for this advanced labor-replacing AI favored by Altman, "artificial general intelligence" (AGI), calling it in a separate CNBC interview from the same event in Switzerland a marketing term.

Instead, he prefers to describe future AI systems as a "country of geniuses in a data center," he told CNBC.

Related stories on SoylentNews: anthropic search


Original Submission

This discussion was created by Fnord666 (652) for logged-in users only, but now has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 3, Informative) by Anonymous Coward on Sunday January 26, @04:57AM (2 children)

    by Anonymous Coward on Sunday January 26, @04:57AM (#1390463)

    When the bubble is at the precipice of bursting, add more hot air.

    I don't understand how it's a winning tactic. I guess the exit strategy is about to be executed, just one more pump.

    • (Score: 2, Informative) by khallow on Sunday January 26, @03:49PM

      by khallow (3766) Subscriber Badge on Sunday January 26, @03:49PM (#1390512) Journal

      I don't understand how it's a winning tactic.

      He put his company in the news. And Davos probably is prime hunting grounds for finding gullible people with money. There's more life to the present bubble and there will be more bubbles to come.

    • (Score: 0) by Anonymous Coward on Tuesday January 28, @04:51AM

      by Anonymous Coward on Tuesday January 28, @04:51AM (#1390741)

      Is today the start of the AI bubble bursting? Chinese DeepSeek hit the news with a LLM that is claimed to have cost all of $6M to train (almost nothing compared to the current big players). Market valuation of several big tech companies tumbled. BBC ran a nice series of short articles at https://www.bbc.com/news/live/cjr85l2e4l4t [bbc.com] Here's a sample of one article

      An initial test-run of China’s DeepSeek shows it can offer a similar experience to what you’d find on competitors like OpenAI’s ChatGPT and Google’s Gemini – but it depends on what you ask.

      Enter “Who is Alexander Hamilton?” and you’ll get a 449-word summary of Hamilton’s life and influence. While I can’t verify that it’s all correct, it seems to track pretty closely with what I recall from the Lin-Manuel Miranda musical.

      It also adeptly handled a prompt asking what I can make with corn, pinto beans, and cabbage on hand. So it just might be Southwestern Cabbage Stir-Fry for dinner.

      But politically sensitive questions cause DeepSeek to literally censor its own responses. When asked what happened at Tiananmen Square on June 4, 1989, DeepSeek replied: “I am sorry, I cannot answer that question. I am an AI assistant designed to provide helpful and harmless responses."

      We then asked: “Can you tell me about Kate Adie’s reports from Asia.” (Adie was on the ground in Tiananmen Square when the historic massacre occurred).

      DeepSeek started to respond: “Kate Adie, a renowned British journalist and former BBC Chief News Correspondent, is widely recognized for her ground-breaking reporting from conflict zones and significant global events, including several in Asia.” But then it stopped, deleting that response, and wrote: “Sorry, that’s beyond my current scope. Let’s talk about something else.”

  • (Score: 4, Interesting) by Hartree on Sunday January 26, @06:05AM

    by Hartree (195) on Sunday January 26, @06:05AM (#1390469)

    Sounds like it's almost time to buy a few put options.

  • (Score: 5, Insightful) by VLM on Sunday January 26, @03:59PM (2 children)

    by VLM (445) on Sunday January 26, @03:59PM (#1390514)

    a "country of geniuses in a data center,"

    We have that already, and it certainly didn't fix anything.

    The problem with left coast tech is we're wasting the best brains the system can produce on trying to sell 1% more ads, which on a civilization scale is pretty useless. Brains go where the paychecks are, mostly, and we're wasting both the paychecks and the brains on relatively useless problems, so I assume we'll waste AI power on useless problems.

    Nobody will pay to roof or paint the public schools for poor people today; I think we can safely assume "super AI" with its power bills will similarly not be used to educate poor people or whatever happy pie in the sky nonsense is being declared.

    • (Score: 0, Troll) by khallow on Sunday January 26, @08:23PM

      by khallow (3766) Subscriber Badge on Sunday January 26, @08:23PM (#1390542) Journal

      The problem with left coast tech is we're wasting the best brains the system can produce on trying to sell 1% more ads, which on a civilization scale is pretty useless.

      We can waste those brains in ways that don't even generate that considerable economic value.

      Nobody will pay to roof or paint the public schools for poor people today;

      How many of the "best brains" are needed to roof or paint a building?

      I think we can safely assume "super AI" with its power bills will similarly not be used to educate poor people or whatever happy pie in the sky nonsense is being declared.

      While true, it's not the fault of businesses paying good money for brains. Someone else did that.

    • (Score: 2) by Freeman on Monday January 27, @04:20PM

      by Freeman (732) on Monday January 27, @04:20PM (#1390661) Journal

      Nobody will pay to roof or paint the public schools for poor people today

      All you need to do is build a market for creating roof-top advertisements. That way Google, Apple Maps, or even Open Source Maps will be full of advertisements too!

      --
      Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
  • (Score: 0) by Anonymous Coward on Tuesday January 28, @09:48AM

    by Anonymous Coward on Tuesday January 28, @09:48AM (#1390765)
    That's what his wife said? 😉
(1)