Google,eroticism and creativity OpenAI, DeepSeek, et al. are nowhere near achieving AGI (Artificial General Intelligence), according to a new benchmark.
The Arc Prize Foundation, a nonprofit that measures AGI progress, has a new benchmark that is stumping the leading AI models. The test, called ARC-AGI-2 is the second edition ARC-AGI benchmark that tests models on general intelligence by challenging them to solve visual puzzles using pattern recognition, context clues, and reasoning.
This Tweet is currently unavailable. It might be loading or has been removed.
According to the ARC-AGI leaderboard, OpenAI's most advanced model o3-low scored 4 percent. Google's Gemini 2.0 Flash and DeepSeek R1 both scored 1.3 percent. Anthropic's most advanced model, Claude 3.7 with an 8K token limit (which refers to the amount of tokens used to process an answer) scored 0.9 percent.
The question of how and when AGI will be achieved remains as heated as ever, with various factions bickering about the timeline or whether it's even possible. Anthropic CEO Dario Amodei said it could take as little as two to three years, and OpenAI CEO Sam Altman said "it's achievable with current hardware." But experts like Gary Marcus and Yann LeCun say the technology isn't there yet and it doesn't take an expert to see how fueling AGI hype is advantageous to AI companies seeking major investments.
The ARC-AGI benchmark is designed to challenge AI models beyond specialized intelligence by avoiding the memorization trap — spewing out PhD-level responses without an understanding of what it means. Instead it focuses on puzzles that are relatively easy for humans to solve because of our innate ability to take in new information and make inferences, thus revealing gaps that can't be resolved by simply feeding AI models more data.
"Intelligence requires the ability to generalize from limited experience and apply knowledge in new, unexpected situations. AI systems are already superhuman in many specific domains (e.g., playing Go and image recognition)" read the announcement.
SEE ALSO: I compared Sesame to ChatGPT voice mode and I'm unnerved"However, these are narrow, specialized capabilities. The 'human-ai gap' reveals what's missing for general intelligence - highly efficiently acquiring new skills."
To get a sense of AI models' current limitations, you can take the ARC-AGI test for yourself. And you might be surprised by its simplicity. There's some critical thinking involved, but the ARC-AGI test wouldn't be out of place next to the New York Timescrossword puzzle, Wordle, or any of the other popular brain teasers. It's challenging but not impossible and the answer is there in the puzzle's logic, which is something the human brain has evolved to interpret.
OpenAI's o3-low model scored 75.7 percent on the first edition of ARC-AGI. By comparison, its 4 percent score on the second edition shows how difficult the test is, but also how there's a lot more work to be done with reaching human level intelligence.
Topics Google OpenAI
Best fitness deal: The Merach R50 rowing machine is 35% off at Amazon
Trump and the media could never get away with fat
Now Facebook can more easily delete your trash Pages
Apple dismisses 200 employees from its Project Titan group
NYT mini crossword answers for January 3, 2025
'The Lego Movie 2' is no 'The Lego Movie': Review
Donald Trump's new DC hotel spray
Instagram now lets you put Story filters on existing photos and videos
How to Get Your Significant Other Into Gaming
Apple changes its mind, says it'll pay photo contest winners after all
Greenpeace activists charged after unfurling 'Resist' banner at Trump Tower in Chicago
50 percent of Facebook users could be fake, report claims
Microsoft's Bing blocked in China
'Bohemian Rhapsody' pulled from GLAAD Media Awards after sexual assault allegations
NYT Connections hints and answers for February 1: Tips to solve 'Connections' #601.
Donald Trump and 'Playboy': A lengthy history
YouTube promises to stop recommending flat Earth and 9/11 truther videos
Elon Musk says ominous music, 'Rick and Morty' will keep Teslas safe
NYT Connections hints and answers for January 20: Tips to solve 'Connections' #589.
I don't like horror. Unless it's in a video game.
YouTube blocked North Korean government’s channelTroll sends strobing GIF to journalist with epilepsy, triggers seizureKFC Japan offers twoRIP Craig Sager: Basketball world mourns an NBA fan favoriteIf you lose one AirPod, you can get a new one for $69Facebook finally cracks down on fake newsAirline skewers vlogger over stowaway video they say is 100% fakeNow you can ask Alexa if what happens in Vegas stays in VegasTroll sends strobing GIF to journalist with epilepsy, triggers seizureAirPods are reportedly coming to Apple's retail stores on Dec. 19'Overwatch,' 'CS:GO' and everything you need to watch in esports this weekend'Rogue One' early box office estimates have cleared the planet'DuckTales' reboot casts David TennantFacebook proves, yet again, that it's terrible at mathVanity Fair is riding high on its Trump feudNasty Women's Choir: The hilarious Christmas carolers we need at the end of 2016'DuckTales' reboot casts David TennantWe're all living in 'Rogue One' nowSo much complaining: 'Super Mario Run' will destroy your data and drain your walletThe 10 best esports moments of 2016 NASA spacecraft spots dead robot on Mars surface BYD ramps up EV push in India with launch of Sealion 7 · TechNode Scientists found an incandescent planet. It's 'constantly exploding.' Solar eclipse 2024: The best internet reactions and memes Dating app fatigue has led to a flurry of IRL singles events Target Circle: 20% off for teachers & college students Voyager spacecraft gave us a scare. But NASA's bringing it back to life. NASA's Voyager is in hostile territory. It's 'dodging bullets.' Pluto's 'heart' is yet another bummer for the dwarf planet Netflix’s 'Supacell' turns stereotypes on their head Wordle today: The answer and hints for July 6 TikTok announces restoration of US services · TechNode What not to do during the imminent 2024 solar eclipse NASA asks: Can anyone help us get our Mars samples back? Apple approves Epic Games Store in Europe, but not without some drama first 'House of the Dragon' Season 2, episode 4: Who is Alyn? New Mars images show the Red Planet's 'Inca City' 'The Sims 4' adds polyamory in its Lovestruck Expansion Pack Wordle today: The answer and hints for July 7 How digital driver's licenses work
3.701s , 8288.015625 kb
Copyright © 2025 Powered by 【eroticism and creativity】,Evergreen Information Network