when

Opus

Advances in artificial intelligence to “human levels” may arrive as soon as 2026, according to Anthropic CEO Dario Amodei.
Speaking to podcaster Lex Fridman this week, Amodei compared the recent advances in artificial general intelligence (<a data-ct-non-breakable="null" href="https://cointelegraph.com/news/openai-agi-readiness-chief-quits" rel="null" target="null" text="null" title="null">AGI</a>) to different levels of education.&nbsp;
“We’re starting to get to PhD level, and last year we were at the undergraduate level, and the year before, the level of a high school student,” he <a data-ct-non-breakable="null" href="https://www.youtube.com/watch?v=ugvHCXCOmm4" rel="null" target="null" text="null" title="null">said</a> before predicting that AI could exceed human capabilities within the next couple years.&nbsp;
<blockquote>
 “If you just eyeball the rate at which these capabilities are increasing, it does make you think that we’ll get there by 2026 or 2027.”&nbsp;
</blockquote>
He added that several things could derail these advancements, such as running out of data, cluster scaling limitations or geopolitical conflicts impacting microchip production.&nbsp;
Still, “if you believe the straight line extrapolation” of current advancement rates, “we’ll get there in 2026 or 2027,” though there might be some “mild delay,” he said.&nbsp;
<figure>
 <img alt="ChatGPT" src="https://s3.cointelegraph.com/uploads/2024-11/01932910-d184-7a01-af99-10fa3125d9bf" title="">
 <figcaption style="text-align: center;">
 Dario Amodei. Source: <a data-ct-non-breakable="null" href="https://www.youtube.com/watch?v=ugvHCXCOmm4" rel="null" target="null" text="null" title="https://www.youtube.com/watch?v=ugvHCXCOmm4">YouTube</a>
 </figcaption>
</figure>
Amodei also expressed concern about the potential implications of human-level AI, saying, “Things that are powerful can do good things, and they can do bad things,” before adding, “With great power comes great responsibility.”&nbsp;
Anthropic’s flagship product is the large language model chatbot Claude, which <a data-ct-non-breakable="null" href="https://cointelegraph.com/learn/articles/elon-musks-grok-ai-chatbot" rel="null" target="null" text="null" title="null">rivals</a> OpenAI’s <a data-ct-non-breakable="null" href="https://cointelegraph.com/news/chatgpt-ai-search-prototype" rel="null" target="null" text="null" title="null">ChatGPT</a>, among others.&nbsp;
When asked about the competition, he reiterated the company’s “race to the top” mission, which involves “trying to push the other players to do the right thing by setting an example.”&nbsp;
Asked when the next iteration of the firm’s <a data-ct-non-breakable="null" href="https://cointelegraph.com/news/how-to-train-an-ai-bot-to-day-trade-crypto" rel="null" target="null" text="null" title="null">AI</a> chatbot, Claude Opus 3.5, was coming out, Amodei was elusive and didn’t offer an exact date.&nbsp;&nbsp;
Related: <a data-ct-non-breakable="null" href="https://cointelegraph.com/news/anthropic-ai-could-one-day-sabotage-humanity-fine-now" rel="null" target="null" text="null" title="null">Anthropic says AI could one day ‘sabotage’ humanity but it’s fine for now</a>
OpenAI CEO <a data-ct-non-breakable="null" href="https://cointelegraph.com/news/sam-altman-chatgpt-unimaginable-prosperity-fix-climate" rel="null" target="null" text="null" title="null">Sam Altman</a> has made similar predictions regarding the arrival of artificial general intelligence within a few years. He recently indicated that the firm was on track to achieve AGI within the next five years with current hardware, <a data-ct-non-breakable="null" href="https://www.msn.com/en-us/money/markets/sam-altman-claims-agi-will-whoosh-by-in-5-years-with-surprisingly-little-societal-change-while-anthropic-ceo-predicts-a-2026-or-2027-breakthrough-theres-no-ceiling-below-the-level-of-humanstheres-a-lot-of-room-at-the-top-for-ais/ar-AA1u2IRk" rel="null" target="null" text="null" title="null">according</a> to reports.&nbsp;
AGI is defined as a type of artificial intelligence that matches or surpasses human capabilities across a wide range of cognitive tasks.&nbsp;
Magazine: <a data-ct-non-breakable="null" href="https://cointelegraph.com/magazine/a-bizarre-cult-is-growing-around-ai-created-memecoin-religions-ai-eye/" rel="null" target="null" text="null" title="null">A bizarre cult is growing around AI-created memecoin ‘religions’: AI Eye</a>

Human-level AI could be here as early as 2026: Anthropic CEO

Cointelegraph

Other articles published on Nov 14, 2024

Meme Coin Whales Are Massively Accumulating PEPE – Will Price Keep Pumping?

Canary Capital Files for a Spot Solana ETF in the U.S.

BlackRock Discloses New Position in IBIT

Bitcoin Spot ETF Inflows Hit $510M, Marking Six-Day Streak

BexBack Exchange Launches Double Deposit Bonus and $50 Welcome Bonus

Market Analysts Predict Solana’s (SOL) Price Surge to New All-Time Highs: Here’s the Outlook

Join our free newsletter for daily crypto updates!