Majarra acquires Arabic AI startup Lableb

Majarra acquires Arabic AI startup Lableb
Lableb will maintain its operational independence under Majarra’s ownership. (Supplied)
Short Url
Updated 19 July 2024
Follow

Majarra acquires Arabic AI startup Lableb

Majarra acquires Arabic AI startup Lableb
  • Startup’s solutions embraced by a diverse clientele

DUBAI: Arabic digital content provider Majarra has announced the acquisition of Lableb, an Arabic language artificial intelligence solutions startup.

The transaction makes Majarra the majority owner of Lableb, enabling a more complete integration of the two entities.

It also aligns with Majarra’s vision to enhance the utility of Arabic online through reliable content, advanced user experiences, and Arabic language technologies.

Lableb’s AI and neuro-linguistic programming technologies power content discovery and personalization software products, including enterprise search and recommendation engines.

The startup’s solutions have been embraced by a diverse clientele, including online stores, government platforms, news websites, and enterprise software providers, where nuanced and accurate Arabic language processing is crucial for success.

Lableb has previously collaborated closely with Microsoft and AWS, in addition to leading e-commerce platforms, including Zid, Salla, and Shopify, and content management system software providers like NVSSoft.

As one of Lableb’s earliest clients, Majarra has firsthand experience of Lableb’s Arabic technologies.

Manhom.com, which is powered by Lableb’s named entity recognition technologies, has become the region’s premier Arabic source for professional information.

Lableb’s innovations have also driven search and content discovery across all Majarra platforms, including the flagship Majarra app.

Lableb will maintain its operational independence under Majarra’s ownership, focusing on addressing the unique challenges machines face with Arabic language processing.

The language’s rich morphology, diverse dialects, complex syntactic structure, and context-dependent meanings have long posed significant hurdles for machine understanding and processing.

Abdulsalam Haykal, Majarra’s executive chairman, and CEO Ammar Haykal said: “Majarra and Lableb share a common foundation and vision.

“Kinda’s (Lableb’s CEO Kinda Altarbouch) leadership and the team’s unwavering commitment to the transformative power of Arabic AI/NLP for regional businesses have consistently delivered exceptional product quality.

“Through Lableb, we see immense potential in bringing real-life applications to large language models, some of which have emerged from our region.”

Altarbouch, who is also Lableb’s co-founder, said: “Joining Majarra marks an exciting new chapter for Lableb. Our shared vision of advancing Arabic AI and NLP will drive innovation and deliver significant value to online businesses and their customers.

“Lableb’s tools are robust, ready to deploy, and capable of handling millions of queries weekly.”

Marcus Brauchli, the managing director of North Base Media, said: “This acquisition is timely and demonstrates Majarra’s agility in responding to market evolution.

“By incorporating AI products into its portfolio, Majarra adds a crucial technology dimension to its offerings. While LLMs are transformative, their true potential can only be realized through AI agents like those Lableb provides.”


Report says crowd-sourced fact checks on X fail to address flood of US election misinformation

Report says crowd-sourced fact checks on X fail to address flood of US election misinformation
Updated 31 October 2024
Follow

Report says crowd-sourced fact checks on X fail to address flood of US election misinformation

Report says crowd-sourced fact checks on X fail to address flood of US election misinformation

SAN FRANCISCO, California: X’s crowd-sourced fact-checking program, called Community Notes, isn’t addressing the flood of US election misinformation on Elon Musk’s social media platform, according to a report published Wednesday by a group that tracks online speech.
The nonprofit Center for Countering Digital Hate analyzed the Community Notes feature and found that accurate notes correcting false and misleading claims about the US elections were not displayed on 209 out of a sample of 283 posts deemed misleading — or 74 percent.
Misleading posts that did not display Community Notes even when they were available included false claims that the 2020 presidential election was stolen and that voting systems are unreliable, CCDH said.

From a page of the "Rated Note Helpful" report of the Center for Countering Digital Hate

In the cases where Community Notes were displayed, the original misleading posts received 13 times more views than their accompanying notes, the group added.
Community Notes lets X users write fact checks on posts after the users are accepted as contributors to the program. The checks are then rated by other users based on their accuracy, sources, how easily they are to understand, and whether they use neutral language. The program was launched in 2021 by the previous leadership of the site — then known as Twitter — and was called Birdwatch. Musk renamed it Community Notes after he took over the site in 2022.

Formerly known as Birdwatch under the old Twitter, the program was renamed Community Notes after Elon Musk took over the popular social media platform in 2022 and renamed it X. (Image courtesy of X)

Last year, X sued CCDH, blaming the group for the loss of “tens of millions of dollars” in advertising revenue after it documented an increase in hate speech on the site. The lawsuit was dismissed by a federal judge in March.
Keith Coleman, a vice president of product at X who oversees Community Notes, said in a statement that the program “maintains a high bar to make notes effective and maintain trust across perspectives, and thousands of election and politics related notes have cleared that bar in 2024. In the last month alone, hundreds of such notes have been shown on thousands of posts and have been seen tens of millions of times. It is because of their quality that notes are so effective.”
San Francisco-based X also pointed to external academic research that has shown Community Notes to be trustworthy and effective.
Imran Ahmed, the CEO of CCDH, however, said the group’s research “suggests that X’s Community Notes are little more than a Band Aid on a torrent of hate and disinformation that undermines our democracy and further polarizes our communities.”


Meta says Malaysia’s social media licencing plan lacks clarity, threatens innovation

Meta says Malaysia’s social media licencing plan lacks clarity, threatens innovation
Updated 30 October 2024
Follow

Meta says Malaysia’s social media licencing plan lacks clarity, threatens innovation

Meta says Malaysia’s social media licencing plan lacks clarity, threatens innovation
  • Malaysia to require social media licenses by Jan. 1, says platforms must step up monitoring
  • Meta says undecided if it will apply for licencing by deadline due to “exceptionally accelerated” iter and unclear obligations

KUALA LUMPUR: A Meta Platforms official on Wednesday criticized Malaysia’s plan to require social media platforms to apply for a regulatory license by January, saying the proposal lacked clear guidelines and gave social companies little time to comply, risking digital innovation and growth in the country.
Malaysia said in July it will require social media platforms and messaging services with more than eight million users to obtain a license, as part of efforts to curb financial scams, cyberbullying and sexual crimes online.
The companies could face legal action if they failed to do so by Jan. 1, 2025.
The plan has faced a backlash, with an Asian industry group that includes Meta urging the government in August to pause the move. Malaysia however has said it will not delay the proposed regulations, with Communications Minister Fahmi Fadzil saying tech companies must comply with local laws to continue operating the country.
Meta’s director of public policy for Southeast Asia Rafael Frankel said the firm has not yet decided if it plans to apply for the license ahead of the January deadline, due to a lack of clarity over the new regulations.
The timeline to apply for a license was “exceptionally accelerated” and the obligations for social media firms under the plan remained unclear, Frankel said in an interview with Reuters.
“These regulations tend to take a couple years to go through multiple iterations... to properly structure them and to balance the need for safety and ensure that you don’t inadvertently cap innovation and digital economic growth,” he said.
Malaysia’s communications ministry did not immediately respond to a request for comment on the interview.
In a Facebook post on Wednesday following a meeting with Meta representatives a day earlier, communications minister Fahmi thanked the company for its willingness to cooperate with the government but urged it to take more proactive measures against sexual content involving minors on its platforms.
Frankel said Meta shared the Malaysian government’s goal for a safe online environment and was working closely with the communications regulator to remove or restrict harmful content from its platforms.
“We don’t need a licencing regime in order to take online safety seriously. We already take it seriously,” he said.
Meta has shared its concerns with the government and hopes to “bridge the differences” over the proposed regulations before they are implemented, Frankel added, without providing details.
Malaysian authorities deem online gambling, scams, child pornography and grooming, cyberbullying and content related to race, religion and royalty as harmful.
Malaysia reported a sharp increase in harmful social media content earlier this year and urged social media firms, including Facebook parent Meta and short video platform TikTok (8645.HK), to step up monitoring on their platforms.


Journalist shot dead in crime-riddled western Mexico

Journalist shot dead in crime-riddled western Mexico
Updated 30 October 2024
Follow

Journalist shot dead in crime-riddled western Mexico

Journalist shot dead in crime-riddled western Mexico
  • Mauricio Cruz Solis was a host on local radio station La Poderosa Uruapan who also published news on the Minuto x Minuto outlet
  • Mexico is one of the world’s most dangerous countries for journalists, news advocacy groups say

MORELIA, Mexico: A journalist was shot dead Tuesday night in western Mexico, a local prosecutor’s office said, in a part of the country hit hard by organized crime.
Mauricio Cruz Solis, a host on local radio station La Poderosa Uruapan who also published news on the Minuto x Minuto outlet, was killed in the city of Uruapan.
One other person was wounded in the attack, the prosecutor’s office said.
The radio station where Cruz Solis worked mourned his killing in a statement published on social media.
“Mauricio was more than a colleague, he was an unconditional friend, a source of inspiration and a tireless voice in the service of our community,” the station said.
Wracked by violence related to drug trafficking, Mexico is one of the world’s most dangerous countries for journalists, news advocacy groups say.
Reporters Without Borders says more than 150 newspeople have been killed in Mexico since 1994.
Cruz Solis’s murder is the first killing of a journalist under the government of Claudia Sheinbaum, who took office on October 1, although there have been other attacks on media this month.
On October 18, gunmen shot at the front of the El Debate newspaper’s office in Culiacan, the state capital of cartel stronghold Sinaloa, which has been shaken by weeks of gang infighting.
A day later, a delivery worker with the outlet was abducted by presumed criminal groups and there has been no news about his whereabouts.


Google reports strong growth driven by AI, cloud

Google reports strong growth driven by AI, cloud
The results showed that Google is holding its own despite worry that the search engine juggernaut has been caught short in the A
Updated 30 October 2024
Follow

Google reports strong growth driven by AI, cloud

Google reports strong growth driven by AI, cloud

San Francisco - USA
San Francisco, Oct 29, 2024 (AFP) -Google's parent company Alphabet reported robust third-quarter results on Tuesday, with revenue climbing 15 percent year-over-year to $88.3 billion, powered by strong performance from its crucial advertising business and growth in cloud services.
The technology giant's net income surged 34 percent to $26.3 billion, while earnings per share increased 37 percent to $2.12.
Operating margins expanded significantly, rising 4.5 percentage points to 32 percent, reflecting improved cost efficiency at one of the world's biggest companies.
"Our commitment to innovation, as well as our long-term focus and investment in AI, are paying off with consumers and partners benefiting from our AI tools," said CEO Sundar Pichai, referring to the burgeoning field of artificial intelligence.
The results showed that Google is holding its own despite worry that the search engine juggernaut has been caught short in the AI race.
Google also faces increasing pressure as its world-dominating search engine competes for ads with a widening array of platforms, including Meta, Amazon, Apple, TikTok and even streamers such as Netflix.
But Emarketer senior analyst Evelyn Mitchell-Wolf said Google proved that its defense "is locked in, and it heads into the holiday season well-positioned to win ad budgets."
The company also satisfied questions about whether its search engine will withstand the growing popularity of generative AI chatbots like ChatGPT, which can answer many queries directly in lieu of an internet search.
In response to these pressures, Google earlier this month reshuffled its search engine business, replacing the division chief after a four-year tenure.
The company also moved its Gemini chatbot team to Google DeepMind, separating it from the search group.
Pichai said that Google Search's AI Overviews, which display a snapshot of information at the top of the results page, was proving a success and available to one billion users in more than 100 countries.
"We are definitely expanding what's possible in search and it's been really heartening to see users adapt," Pichai told analysts.
Google Cloud emerged as a standout performer in the June-to-September period, with revenue jumping 35 percent to $11.4 billion, driven by increased adoption of AI infrastructure and solutions.
That business, which trails market leaders Amazon and Microsoft, provides businesses with computing power, data storage and AI tools delivered over the internet.
The company's core Google Services segment, which includes Search, YouTube, and other advertising products, grew 13 percent to $76.5 billion.
YouTube hit a significant milestone, with total advertising and subscription revenues exceeding $50 billion over the past four quarters for the first time.
After the release, investors sent Google's share price higher by five percent in after-hours trading.
Despite the solid earnings, Google faces significant legal challenges, at least in the long-term.
In August, a federal judge found that Google's search engine was an illegal monopoly, with the US government suggesting a company breakup might be appropriate.
A separate antitrust lawsuit challenges its dominance of online advertising.
Additionally, Google has been ordered to open Android-powered smartphones to rival app shops following a case brought by Fortnite-maker Epic Games, though that decision has been suspended pending appeal.
While those legal battles are expected to take years to resolve, they have raised questions about the company's long-term strategy and its heavy reliance on search engine advertising revenue.


CNN issues apology after panelist’s ‘terrorist’ remark directed at Mehdi Hasan

CNN issues apology after panelist’s ‘terrorist’ remark directed at Mehdi Hasan
Updated 29 October 2024
Follow

CNN issues apology after panelist’s ‘terrorist’ remark directed at Mehdi Hasan

CNN issues apology after panelist’s ‘terrorist’ remark directed at Mehdi Hasan
  • Political commentator Ryan James Girdusky said he hoped Hasan’s ‘beeper doesn’t go off,’ in apparent reference to Israel’s pagers attack
  • ‘There is zero room for racism or bigotry at CNN,’ network said

LONDON: CNN issued an apology after a panelist on its program suggested that fellow guest Mehdi Hasan was a “terrorist.”

The incident occurred on Monday night’s broadcast of NewsNight, when conservative commentator Ryan James Girdusky told Hasan, the founder of the recently launched news site Zeteo and former MSNBC host, that he hoped his “beeper doesn’t go off.”

This appeared to reference Israel’s recent targeting of Hezbollah fighters in Lebanon using explosive pager devices, which killed 12 and injured many more.

“Did your guest just say I should be killed on live TV?” Hasan asked the show’s anchor, Abby Phillip.

After a commercial break, Phillip issued an on-air apology to Hasan and viewers and said Girdusky had been removed from the show.

“I want to apologize to Mehdi Hasan for what was said at this table. It was completely unacceptable,” she said. “I want to apologize to the viewers at home.”

Hasan also did not reappear on the show after the commercial break, leading to some speculation that he was also asked to leave, although Philip later said in a video on X, which was also reposted by Hasan, that the network “really wanted him to come back and finish the show.”

The exchange began when the two guests were invited to discuss the upcoming US election and ongoing conflicts in the Middle East.

Hasan referenced Donald Trump’s recent rally and accusations that the Trump campaign is spreading Nazi-style ideas.

Girdusky then accused Hasan, an outspoken advocate for Palestinian rights, of being “the most antisemitic person at this table,” a label Hasan said he frequently faces due to his views.

“Yeah, well, I hope your beeper doesn’t go off,” Girdusky then said, before attempting to justify his words by claiming he thought Hasan said he supported Hamas.

“You didn’t think I said Hamas. I said I’m a supporter of Palestinian rights,” Hasan said, adding, “At least have the guts to support your racist comment.”

In a later post on X, however, the conservative political consultant and writer appeared to double down on his remarks, saying, “You can stay on CNN if you falsely call every Republican a Nazi and have taken money from Qatar-funded media,” he said. “Apparently you can’t go on CNN if you make a joke. I’m glad America gets to see what CNN stands for.”

Following the exchange, CNN announced that it would no longer feature appearances from Girdusky.

“There is zero room for racism or bigotry at CNN or on our air,” the network said in a statement. “We aim to foster thoughtful conversations and debate including between people who profoundly disagree with each other in order to explore important issues and promote mutual understanding.

“But we will not allow guests to be demeaned or for the line of civility to be crossed. Ryan Girdusky will not be welcomed back at our network.”

The incident prompted widespread condemnation, with several people describing the comments as an overtly racist attack.

Congresswoman Rashida Tlaib, who is Muslim and Palestinian-American, wrote on X: “CNN has played a central role in normalizing this kind of open anti-Arab, anti-Muslim, and anti-Palestinian hate, and it’s sadly totally unsurprising to see such disgusting statements aired so casually on its network now.”