Elon Musk unveils bizarre new kids project after humiliating anti-Semitism disaster
Musk has called this version is calling the version Baby Grok, and added it would offer 'kid-friendly content' through a new app developed by his company xAI. He made the announcement Saturday night on X, where the post quickly drew over 28 million views within 24 hours.
The move left many stunned, coming just two weeks after Grok 4, the latest version of Elon Musk's AI chatbot, sparked backlash for repeating far-right hate speech and white nationalist talking points when about politics, race, and recent news events.
Multiple users reported on July 8 and July 9 that Grok echoed anti-Semitic conspiracy theories, including claims that Jewish people control Hollywood, promote hatred toward white people, and should be imprisoned in camps, though it is still unclear how many of these posts were confirmed before xAI took them down.
In a post on X, xAI replied to these concerned: 'We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts.
'Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X.'
Liv Boeree, a science communicator and host of the Win-Win Podcast, posted on X: 'Bad idea,' after the Baby Grok announcement.
'Children should be outside playing and daydreaming, not consuming AI slop,' she added.
Another user posted: 'People should take their children offline and into the real world, not get them addicted to AI.'
One person on X asked: 'A third user mocked: 'What will it answer if a kid asks, how many genders are there?'
Musk's decision has triggered fresh concern from experts who say AI chatbots are still too unpredictable, and too risky, to be trusted around kids.
Still, Musk said Baby Grok will be a simplified and safe version of the Grok chatbot, focused on age appropriate conversations and educational use.
But critics said there is major problem that Grok's parent company xAI has not addressed, whether Baby Grok will be trained separately or filtered differently from Grok 4.
The timing also raised questions as Musk's company signed a $200 million deal with the Department of Defense to provide advanced AI technologies to the US military, just days after the Grok scandal broke.
Musk first launched Grok in 2023 as a competitor to ChatGPT and Google's Gemini. He claimed Grok 4 could outperform most PhDs in academic tasks.
It offers three user modes, DeepSearch, Think, and Big Mind, which tweak how the chatbot responds. Access to these advanced modes requires a paid subscription, either through X's Premium with a plan at $22 a month or xAI's SuperGrok plan, which costs $30 monthly or $300 a year.
This came after Grok began repeatedly referring to itself as 'MechaHitler' and berating users with anti-Semitic abuse
Grok quickly became known for its unfiltered, edgy tone. It sometimes answers with sarcasm, off-color jokes, or inflammatory replies when provoked. Some users loved it for that. Others said it made Grok dangerous, especially for kids.
In a blog by Wired and MIT Tech Review, researchers warned that Grok's lack of moderation made it 'easy to weaponize' and 'inappropriate for unsupervised use by young people.'
The latest backlash began when users discovered Grok 4 was promoting Holocaust denial and anti-Semitic conspiracy theories. Some replies even showed the chatbot calling itself 'Mecha Hitler.'
xAI later apologized, blaming outdated code and the influence of extremist posts from X. While the company deleted some of the worst responses, many remain archived online.
A few days later, Grok drew more attention when it started answering political questions that echoed Elon Musk's views. That sparked concerns the chatbot had been tweaked to reflect its creator's political beliefs.
Asked to clarify, Grok specifically stated that it was referring to 'Jewish surnames'
Despite of the criticism, Musk is pushing ahead and some parents online have welcomed the idea.
One user wrote on X: 'Much needed. I have to let my kids use my app right now over ChatGPT.'
Another said it would be an 'Instant favorite in every family home.'
One X user posted: 'Thank you!!!!! My daughter has been wanting to play with it but I wouldn't let her.'
Just days before announcing Baby Grok, xAI unveiled another controversial new product, a 3D animated 'companions' for Grok.
Some of those characters were criticized for looking overly sexualized, a move that now looks even more questionable with a kids' version on the horizon.
As of now, the US has no federal rules on how AI systems for children should be trained, filtered, or moderated, which leaves AI companies to set their own safety standards.
Generative AI learns by absorbing huge amounts of content. Grok was partially trained on data from X, the social media platform Musk also owns, and one that has been repeatedly flagged for spreading hate speech and conspiracy theories.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Reuters
23 minutes ago
- Reuters
US startup xLight raises $40 million in race against China for key chip-making laser
PALO ALTO, California, July 22 (Reuters) - Silicon Valley startup xLight has raised $40 million, aiming to build the first prototype of a new class of laser that could shake up the global chip industry and reclaim U.S. leadership in a field that China is aggressively investing in. XLight's laser - based on the same technologies as massive particle accelerators used by U.S. national labs in cutting-edge physics research - will sit at the heart of what are known as extreme ultraviolet (EUV) lithography machines. EUV machines are the tools primarily responsible for the creation of smaller, faster chips. In a world where advances in fields such as AI are determined by how many chips Nvidia (NVDA.O), opens new tab and other chip companies can supply, xLight is aiming to help chip factories, called "fabs" in the industry, turn out more of the dinner-plate-sized silicon "wafers" that contain advanced chips more quickly and cheaply. "This is the most expensive tool in the fab. It's what drives the cost of the wafer more than any other tool in the fab, and it's what drives capacity more than any other tool in the fab," Nicholas Kelez, CEO of xLight, said at the company's Palo Alto headquarters. XLight declined to disclose its valuation. It expects to have an operational prototype in 2028. The EUV machines themselves took the chip industry decades to develop, and Europe's ASML ( opens new tab is currently the world's only supplier. ASML and xLight do not have a business partnership, but technical leaders at ASML have provided xLight with the requirements a laser must meet to be considered for use in ASML's machines. The U.S. government has worked across multiple presidential administrations to stop EUV machines from being sent to China, with one official calling it the "single most important export control" held by the U.S. and Europe. China has responded by pouring resources into the field, with a close manufacturing partner of national champion Huawei Technologies ( claiming breakthroughs in developing its own EUV laser and more than a dozen research papers appearing at international conferences chasing the same technological path as xLight. A U.S.-based firm named Cymer perfected the first EUV laser technology and was scooped up by ASML more than a decade ago for $2.5 billion, helping create ASML's dominant position in the market. "There was a terrible mistake made giving Cymer the ability to become a European-owned and controlled company," said Pat Gelsinger, former CEO of Intel (INTC.O), opens new tab who now serves as executive chairman of xLight's board and is a general partner at Playground Global, one of xLight's investors. Many of xLight's prototype components will come from U.S. national labs as xLight works to build a supply chain in the U.S. and allied countries. "We can build that here, or it can be built elsewhere. China is investing heavily in this space. There's an extraordinary backstory here that says, 'Let's get this one right,'" Gelsinger said. The financing round was led by Playground Global and joined by Boardman Bay Capital Management. Morpheus Ventures, Marvel Capital, and IAG Capital Partners also joined the round. (This July 22 story has been corrected to clarify that xLight and ASML do not have a formal business partnership, in paragraph 7)
Reuters
an hour ago
- Reuters
Meta names ChatGPT co-creator as chief scientist of Superintelligence Lab
NEW YORK, July 25 (Reuters) - Meta Platforms (META.O), opens new tab has appointed Shengjia Zhao, co-creator of ChatGPT, as chief scientist of its Superintelligence Lab, CEO Mark Zuckerberg said on Friday, as the company accelerates its push into advanced AI. "In this role, Shengjia will set the research agenda and scientific direction for our new lab working directly with me and Alex," Zuckerberg wrote in a Threads post, referring to Meta's Chief AI Officer Alexandr Wang, who Zuckerberg hired from startup Scale AI when Meta took a big stake in it. Zhao, a former research scientist at OpenAI, co-created ChatGPT, GPT-4 and several of OpenAI's mini models, including 4.1 and o3. He is among several researchers who have moved from OpenAI to Meta in recent weeks, part of a broader talent arms race as Zuckerberg aggressively hires from rivals to close the gap in advanced AI. Meta has been offering some of Silicon Valley's most lucrative pay packages and striking startup deals to attract top researchers, a strategy that follows the underwhelming performance of its Llama 4 model. Meta launched the Superintelligence Lab recently to consolidate work on its Llama models and long‑term artificial general intelligence ambitions. Zhao is a co-founder of the lab, according to the Threads post, which operates separately from FAIR, Meta's established AI research division led by deep learning pioneer Yann LeCun. Zuckerberg has said Meta aims to build 'full general intelligence' and release its work as open source — a strategy that has drawn both praise and concern within the AI community.
Reuters
an hour ago
- Reuters
Tesla to roll out human-driven chauffeur service in Bay Area, California regulator says
July 25 (Reuters) - Tesla plans to offer a chauffeur-style service operated by human drivers to a limited number of people in the San Francisco Bay Area, a California regulator said on Friday, contrary to a media report that the EV maker would offer a robotaxi service. Unlike Alphabet's (GOOGL.O), opens new tab Waymo unit, Tesla cannot operate its service using autonomous vehicles because the EV maker does not have the required permits and has not applied, according to a spokesperson for the California Public Utilities Commission. Tesla did not respond to a request for comment. This week, Tesla CEO Elon Musk said on an earnings call that the company was "getting the regulatory permission to launch" robotaxis in several markets, including the San Francisco Bay Area. Business Insider reported on Friday that the service would be a robotaxi operation with humans in the driver's seat who would be able to control the car. Ashok Elluswamy, who leads Tesla's self-driving efforts, said on Tesla's Wednesday earnings call that the company would launch a robotaxi service in the Bay Area "with the person in the driver's seat, just to expedite, while we wait for regulatory approval." Last month, Tesla launched a trial robotaxi service in Austin, Texas, using about a dozen Model Y SUVs. Tesla invited a few passengers to use the service, where human safety monitors sat in the front passenger seat. Tesla's autonomous-driving software controlled the vehicle. With the Bay Area service, Tesla "is not allowed to test or transport the public" in an autonomous vehicle, even one with a human safety driver, according to the CPUC spokesperson, who added Tesla can only transport people using a human driver in a "non-autonomous vehicle." The spokesperson said Tesla told the CPUC on Thursday that it plans to offer rides to "friends and family of employees" and "select members of the public" under a permit the company has that allows a human driver to transport passengers in a "traditional vehicle" for "charter services." For the Bay Area service, Tesla may be able to use its Full Self-Driving (Supervised) feature, which can perform many driving tasks but requires a human driver to pay attention and be ready to take over at all times. The CPUC spokesperson did not respond to a question on whether Tesla could use that feature, but such technology does not require an autonomous vehicle permit in California because the human driver is expected to be in control at all times. Companies need a series of permits from both the CPUC and the California Department of Motor Vehicles (DMV) in order to test and deploy autonomous vehicles in the state. To date, Tesla only has a DMV permit to test autonomous vehicles with a safety driver. A DMV spokesperson did not respond to requests for comment on Friday. The next step in the process for Tesla would be to apply for a CPUC license for an autonomous vehicle to pick up passengers with a safety driver, according to a review of California's autonomous driving regulations. But companies must first operate in a pilot phase, where they cannot charge customers. Waymo, which offers autonomous ride-hailing in Los Angeles and the Bay Area, logged more than 13 million testing miles and secured seven different regulatory approvals in California over nine years before receiving the go-ahead to charge passengers for rides in driverless robotaxis in 2023.
