Posted on Leave a comment

Can AI Score As High As A Human On A Test For “General Intelligence”? – IFLScience

CLOSE
We have emailed you a PDF version of the article you requested.
Please check your spam or junk folder
You can also addnewsletters@iflscience.comto your safe senders list to ensure you never miss a message from us.
CLOSE
Complete the form below and we will email you a PDF version
GET PDF
Cancel and go back
IFLScience needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time.
For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, check out ourPrivacy Policy
CLOSE
Complete the form below to listen to the audio version of this article
Listen
Cancel and go back
IFLScience needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time.
For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, check out ourPrivacy Policy
Advertisement
Search
Subscribe today for our Weekly Newsletter in your inbox!
Subscribe today for our Weekly Newsletter in your inbox!
Rosie McCall
Rosie McCall
Freelance Writer
Rosie is a freelance writer living in London. She has covered everything from ancient Egyptian temples to exciting medical breakthroughs, but she particularly enjoys writing about wildlife, anthropology and the wonders of the human mind.
BookView full profile
BookRead IFLScience Editorial Policy
Freelance Writer
Katy Evans
Managing Editor
Katy is Managing Editor at IFLScience where she oversees editorial content from News articles to Features, and even occasionally writes some.
BookView full profile
BookRead IFLScience Editorial Policy
DOWNLOAD PDF VERSION
OpenAI’s latest software scored 82.8 percent on a test for artificial general intelligence. That puts it in league with the average human, say researchers. 
Image credit: Krot_Studio/Shutterstock.com
DOWNLOAD PDF VERSION
AI has smashed records on a program designed to test “general intelligence”, achieving a score on a level with those of the average person. 
Historically, researchers have looked to the Turing Test to measure machine intelligence. To pass, a machine must convince a human that it too is a person. By some accounts, technology has already accomplished this feat. Indeed, ChatGPT may have cracked the test earlier this year. However, scientists question whether this can determine true intelligence. 
As an alternative, software engineer and AI researcher Francois Chollet created the ARC-AGI benchmark test, software designed to measure “artificial general intelligence” (or AGI). According to Chollet, “AGI is a system that can efficiently acquire new skills outside of its training data.” 
On this measure, ChatGPT would fail. The technology relies on probability and vast amounts of data to predict the most likely series of words to any given output. It is extraordinarily talented at creating content. However, Chollet would argue that true general intelligence is not so much about the skill (in this case, generating content) but its ability to acquire that skill in the first place without a huge amount of input. This is an ability ChatGPT lacks.

Therefore, to pass the ARC-AGI benchmark test, AI must complete a series of reasoning problems based on colored squares in a grid. Its task is to identify the pattern that turns one grid into another grid and it is given just three examples to learn from. The previous record (held by Jeremy Berman) was 58.5 percent. That record was smashed by OpenAI’s new o3 system, which scored an impressive 82.8 percent – and arguably puts it in league with humans, Chollet says.
In a blog piece, Chollet describes it as “a significant leap forward” representing a “genuine breakthrough in adaptability and generalization”. He said, “This is not just incremental progress; it is new territory, and it demands serious scientific attention.”
To put it in some context, four years ago, GPT-3 scored a less-than-impressive 0 percent. In 2024, GPT-4o did not do much better at 5 percent. Needless to say, there has been a dramatic rate of improvement. Still, there is no need to get too hasty. As Chollet himself points out, the o3 system still performs badly on some simple tasks.
While there have been some impressive developments when it comes to AI, there is little consensus amongst AI researchers on when we should expect to see true AGI. Some believe it is something we could see by the end of the decade. In a recent talk, Ben Goertzel, founder of SingularityNET, argued individual computers would have power equivalent to a human brain by 2023. “Then you add another 10/15 years on that, an individual computer would have roughly the compute power of all of human society.”
artificial intelligence,
turing test
link to article
link to article
link to article
Advertisement
Advertisement
Advertisement
link to article
link to article
link to article
Sign up today to get weekly science coverage direct to your inbox
&copy 2024 IFLScience. All Rights Reserved. RSS

source

Posted on Leave a comment

Hamster Kombat to debut layer-2 network on TON after failed airdrop – crypto.news

Share
Hamster Kombat, once a top Telegram game, is turning to blockchain development on TON after struggling with declining popularity.
Hamster Kombat, a tap-to-earn game built on Telegram, is showing signs to deepen its blockchain roots, this time launching a layer-2 network on top of The Open Network, or just TON (TON). The decision comes after a community vote favouring the development of a dedicated network seemingly designed to enhance the Hamster Kombat token’s HMSTR (HMSTR) use cases.
💫HERE COMES THE BLOCKCHAIN 💫

The vote on the second proposal has ended! The DAO community has spoken: there WILL BE Hamster L2 blockchain built on TON!

We will be serving our community, the biggest web3 community in the world, with the technological foundation it asked for!… pic.twitter.com/v4p3jydAPF
Table of Contents
Layer-2 networks are built on top of existing blockchains to improve transaction speeds and scalability. While Ethereum already hosts several well-known layer-2 networks, such as Coinbase’s Base or Blast, TON is still in the early stages of exploring similar technologies.
As of press time, there are no indications that TON hosts any operational layer-2 network. However, the TON ecosystem seems to be working on compatibility with Ethereum Virtual Machine protocols. The strategy may be aimed at attracting liquidity and dApp developers from Ethereum to integrate them into TON’s infrastructure.
For instance, in July, another project, dubbed TON Applications Chain, announced plans to develop a layer-2 network on TON. However, TAC aims to leverage Polygon’s layer-2 technology to achieve this goal.
The so-called “Hamster Blockchain” comes from a vote within the Hamster DAO. The proposal suggested building a network “where HMSTR token would be used for gas fees,” per the proposal’s description.
The vote met the required quorum of 1 billion HMSTR tokens on Friday, Dec. 27, with more than 1.18 billion HMSTR tokens in favor and just over 104,000 tokens against. However, it seems that over 60% of the total votes cast in favor of the initiative — amounting to approximately 737.3 million HMSTR tokens — were submitted by a single address, identified as UQDz…hUCu.
Launched in March, Hamster Kombat rapidly gained popularity, amassing millions of users within five months.
According to the Hamster Kombat team, a whopping 300 million people have signed up to play the game. Of these, a little under half of the players — 131 million in total — qualified to receive HMSTR tokens. Meanwhile, over 2.3 million players were flagged as cheaters and excluded from the token distribution.
The game’s integration with Telegram received a boost from the messaging app’s founder, Pavel Durov, who praised the game’s launch on TON as a way to bring blockchain benefits to millions.
However, for Hamster Kombat, things went south pretty quickly following the airdrop of its token, as it failed to meet players’ financial expectations. By early November, the user base had plummeted by 86%, dropping from 300 million to approximately 41 million active monthly players, per data from IntoTheBlock.
Several factors may have contributed to this decline:
In trying to solve these issues, the Hamster Kombat development team appears to be shifting the community’s focus toward the creation of a layer-2 network on TON. Nonetheless, several critical details remain unclear. For example, there’s no information thus far on who will develop the network, what technological stack will be used, or, most importantly, who’s gonna pay for the ambitious plan.
The missing details in the initiative’s description leave too many questions unanswered as well. On top of that, the proposed layer-2 network isn’t mentioned in Hamster Kombat’s roadmap at all, which appears to have not been updated since September.
Read more about
Deep Dives
Hamster Kombat to debut layer-2 network on TON after failed airdrop
Crypto tech projects shouldn’t settle for shallow marketing | Opinion
From physical rewards to solar power: How DePIN works
Texas, Ohio, and Pennsylvania to create the state-level Bitcoin reserves. What are these bills?
Why meme coins are fading this cycle: A tale of Bitcoin and the others
Should web3 devs be building tribes, not just tools? | Opinion
Coinbase CLO shares data on crypto hedge funds debanking, demands for answers
BTC crosses the hallmark of 19.8 mined units on Xmas eve. It’s time to talk about Bitcoin’s supply cap
Bitcoin or altcoins? Find out who will win the holiday season
Crypto taxes: What does regulation look like around the world?
Related News
Ethereum can speed up by running nearly 65% of its transactions in parallel, Sei says
Binance to support Optimism network upgrade and hard fork
Pig butchering crypto scams wiped out $3.6b this year: report
Get crypto market analysis and curated news delivered right to your inbox every week.
You have successfully joined our subscriber list.

source

Posted on Leave a comment

How To Watch, Betting Odds: Ole Miss Basketball Faces Memphis Tigers – Sports Illustrated

The Ole Miss Rebels have enjoyed a solid start to non-conference play this season, currently holding an 11-1 record with one game left before opening the SEC portion of the schedule.
That one game comes on Saturday when the Rebels travel to take on the Memphis Tigers. Memphis comes into this matchup with a 9-3 record with losses to Auburn, Arkansas State and, most recently, Mississippi State on its docket.
Last season, Ole Miss claimed an 80-77 win over Memphis at the SJB Pavilion in Oxford. This annual meeting has evolved into a rather competitive Mid-South rivalry on the hardwood, and Saturday should be another entertaining installment in this series.
Here's how you can watch Ole Miss face Memphis on Saturday as well as the current betting odds for the game.
WHO: No. 16 Ole Miss Rebels vs. Memphis Tigers
RECORDS: Ole Miss (11-1) vs. Memphis (9-3)
WHEN: Dec. 28 — 1 p.m. CT
WHERE: FedEx Forum — Memphis, Tenn.
TELEVISION/STREAMING: ESPN2
Play-by-Play: John Schriffen, Color: Daymeon Fishback
RADIO: Ole Miss Radio Network
Play-by-Play: David Kellum, Color: Marc Dukes
SPREAD (via FanDuel): Ole Miss -1.5, Memphis +1.5
MONEY LINE: Ole Miss -113, Memphis -106
OVER/UNDER: 152.5
Game odds refresh periodically and are subject to change.
If you or someone you know has a gambling problem and wants help, call 1-800-GAMBLER.
John Macon Gillespie is the publisher of The Grove Report and has experience on the Ole Miss beat spanning five years.
Follow JMakeGillespie
© 2024 ABG-SI LLC – SPORTS ILLUSTRATED IS A REGISTERED TRADEMARK OF ABG-SI LLC. – All Rights Reserved. The content on this site is for entertainment and educational purposes only. Betting and gambling content is intended for individuals 21+ and is based on individual commentators' opinions and not that of Sports Illustrated or its affiliates, licensees and related brands. All picks and predictions are suggestions only and not a guarantee of success or profit. If you or someone you know has a gambling problem, crisis counseling and referral services can be accessed by calling 1-800-GAMBLER.

source

Posted on Leave a comment

Internet Archive restores several services after website defacement, DDoS attack – The Record from Recorded Future News


Leadership
Cybercrime
Nation-state
Elections
Technology
Cyber Daily®
Click Here Podcast
✉️ Free Newsletter
The Internet Archive has restored many of its services more than a week after hackers targeted the popular platform with a distributed denial-of-service (DDoS) attack and defaced its website
In an update on Thursday evening, founder Brewster Kahle said that after the incidents last week, the nonprofit was forced to shut down the site in an effort to improve security. 
“The stored data of the Internet Archive is safe and we are working on resuming services safely. This new reality requires heightened attention to cyber security and we are responding. We apologize for the impact of these library services being unavailable,” Kahle said. 
“The Wayback Machine, Archive-It, scanning, and national library crawls have resumed, as well as email, blog, helpdesk, and social media communications. Our team is working around the clock across time zones to bring other services back online.” 
Kahle added that more services will resume in the coming days, with several in read-only mode because full restoration will take more time. The goal, according to Kahle, is to take a “cautious, deliberate approach to rebuild and strengthen” the site’s defenses. 
“Our priority is ensuring the Internet Archive comes online stronger and more secure,” he said. 
The attack emerged on October 10 when a hacker claimed they stole data on 31 million users of the platform that included usernames, emails and encrypted passwords. The Internet Archive disabled the source of the breach, scrubbed systems and upgraded security, according to Kahle.
Privacy expert Troy Hunt confirmed that the hackers did steal real information and added the leak to his HaveIBeenPwnd service, where people can see whether their login information for a service or website has been exposed.
The junk web traffic from DDoS attacks continued for several days and was  eventually claimed by a group of hackers going by the name SN_BLACKMETA. Researchers have noted that while most of its posts are written in Russian and the working hours align with Moscow time, the group has explicitly targeted institutions across the Middle East with powerful DDoS attacks.
The hackers claimed they targeted the Internet Archive “because the archive belongs to the USA, and as we all know, this horrendous and hypocritical government supports the genocide that is being carried out by the terrorist state of ‘Israel.’”
This week, the Justice Department arrested two men who sold a tool that helped groups like SN_BLACKMETA launch DDoS attacks. 
Kahle compared the attack on Internet Archive to several other high-profile cyberattacks on libraries around the world, including ransomware attacks on the British Library, Toronto Public Library, Seattle Public Library and this week’s incident involving the Calgary Public Library
“We hope these attacks are not indicative of a trend,” Kahle explained. 
Jonathan Greig
is a Breaking News Reporter at Recorded Future News. Jonathan has worked across the globe as a journalist since 2014. Before moving back to New York City, he worked for news outlets in South Africa, Jordan and Cambodia. He previously covered cybersecurity at ZDNet and TechRepublic.
Privacy
About
Contact Us
© Copyright 2024 | The Record from Recorded Future News

source