As anticipated based on updates and new settings in the mobile app for Elon Musk’s social network X, a new large language model (LLM) called Grok-2 from Musk’s sister company xAI landed last night — and it’s a doozy.
Integrated within X itself and available through the Premium (USD 7/month) and Premium+ ($14/month with no ads) subscription tiers, Grok-2 comes, fittingly, in two model sizes: Grok-2 and Grok-2 mini. Grok-2 offers state-of-the-art performance in a wide range of tasks including chat, coding, reasoning, and vision-based application, while Grok-2 mini is a smaller, faster version optimized for efficiency, suitable for simpler text-based prompts requiring quicker responses.
Grok-2 not only boasts image generation capabilities based on a partnership with Black Forest Labs and its new and surprisingly photorealistic open-source diffusion AI model Flux.1, but it also shockingly outperforms the AI models from leading rivals including OpenAI (GPT-4o) and Anthropic (Claude 3.5 Sonnet), and even Google (Gemini Pro 1.5) on leading third-party benchmark tests.
A new, surprising leader across multiple benchmarks
Specifically, Grok-2 and Grok-2 mini outperform all other models on the GPQA, MMLU, MMLU-Pro, MATH, HumanEval, MMMU, MathVista and DocVQA benchmarks.
Even the less-chatbot arena, where many companies covertly test their AI models under alternate names in advance of release (including xAI, where Grok-2 was initially called “sus-column-r”) congratulated xAI on the milestone.
As AI influencer and University of Pennsylvania Wharton School of Business professor Ethan Mollick observed on X, “There are now five GPT-4 class models: GPT-4o, Claude 3.5, Gemini 1.5, Llama 3.1 and now Grok 2.”
Musk congratulated his “hardworking xAI team!” on the similarly named social network.
Image generations steal the show
Even though Grok-2 boasts leading performance on all these different benchmarks related to math, writing, code, and other tasks, by far, the marquee feature capturing the most attention from the jump is its integration with Black Forest Labs’ Flux.1 image generation model.
Before the release of Grok-2, Flux.1 had already been making waves in AI and AI art circles more specifically in the last few weeks as people discovered that they could achieve incredibly photorealistic generations from the open source model, enough to resemble familiar situations like a speaker at a TED talk, as well as adapt the model using low-rank adaptation (LoRA) to generate their own likeness in different situations.
Now that a version of Flux.1 is integrated directly into Grok-2 much in the same way OpenAI integrated its image generation model DALL-E 3 directly into ChatGPT, allowing users to simply type text prompts to the chatbot and ask it to make their images on command, users are testing this capability out in Grok-2 and finding it is notably permissive — generating controversial, compromising images even of public figures such as U.S. presidential candidates Kamala Harris and Donald Trump.
Other leading image generators including Midjourney DALL-E 3 and Microsoft Designer have prohibitions around generating this type of content — especially in the wake of the controversy earlier this year over unauthorized explicit deepfakes of popular musician Taylor Swift (made by prompt engineering around the Designer restrictions) — so it is notable that Grok-2 is bucking that trend and allowing for more freedom, and potential risk. However, that is in keeping with Musk’s stated “free speech” ethos for X.
Yet users are raising concerns about what the capability means for the providence of deepfakes and misinformation across the web.
As user @Omiron33 put it well: “Yes, we’ve had MJ and Flux, but this is the first to make it usable and quick. Advertising, Propaganda, and everything good or bad that comes with that just happened (IMO, the good outweighs the bad)”
Elon Musk’s AI company xAI released its latest language model, Grok 2, on Tuesday, introducing powerful image generation capabilities that have flooded X.com (formerly known as Twitter) with controversial content.
Within hours of its launch, X.com users reported a deluge of AI-generated images depicting graphic violence, explicit sexual content, and manipulated photos of public figures in offensive situations.
The rapid proliferation of controversial content on X.com aligns with the platform’s well-known laissez-faire approach to content moderation. It also marks a significant departure from the cautious strategies adopted by other leading AI companies.
Google, OpenAI, Meta, and Anthropic have implemented strict content filters and ethical guidelines in their image-generation models to prevent the creation of harmful or offensive material.
Grok 2’s unrestricted image generation capabilities, on the other hand, reflect Musk’s long-standing opposition to stringent content moderation on social media platforms.
By allowing Grok 2 to produce potentially offensive images without apparent safeguards, xAI has reignited the debate over tech companies’ role in policing their own technologies. This hands-off approach stands in stark contrast to the industry’s recent focus on responsible AI development and deployment.
The release of Grok 2 comes just six months after Google struggled with its own AI image generator. Google’s Gemini AI faced criticism for being overly “woke” in its image generation, producing historically inaccurate and bizarrely diverse images in response to user prompts.
Google admitted that its efforts to ensure diversity “failed to account for cases that should clearly not show a range” and that its AI model became “way more cautious” over time, refusing to answer even innocuous prompts.
Google’s senior vice president Prabhakar Raghavan explained, “These two things led the model to overcompensate in some cases, and be over-conservative in others, leading to images that were embarrassing and wrong.” As a result, Google temporarily paused Gemini’s image generation feature for people while it worked on improvements.
Grok 2, on the other hand, appears to have no such restrictions, aligning with Musk’s long-standing opposition to content moderation on social media platforms.
By allowing Grok 2 to produce potentially offensive images without apparent safeguards, xAI has launched a new chapter in the debate over tech companies’ role in policing their own technologies.
The ethics tightrope: Balancing innovation and responsibility in AI
The AI research community has reacted with a mix of fascination and alarm. While Grok 2’s technical capabilities are impressive, the lack of adequate safeguards raises serious ethical concerns.
The incident highlights the challenges of balancing rapid technological advancement with responsible development and the potential consequences of prioritizing unrestricted AI capabilities over safety measures.
For enterprise technical decision-makers, the Grok 2 release and its aftermath carry significant implications. The incident underscores the critical importance of robust AI governance frameworks within organizations. As AI tools become more powerful and accessible, companies must carefully consider the ethical implications and potential risks associated with deploying these technologies.
The Grok 2 situation serves as a cautionary tale for businesses considering the integration of advanced AI models into their operations. It highlights the need for comprehensive risk assessment, strong ethical guidelines, and robust content moderation strategies when implementing AI solutions, particularly those with generative capabilities. Failure to address these concerns could lead to reputational damage, legal liabilities, and erosion of customer trust.
The ripple effect: Grok 2’s impact on AI governance and social media
Moreover, the incident may accelerate regulatory scrutiny of AI technologies, potentially leading to new compliance requirements for businesses using AI.
Technical leaders should closely monitor these developments and be prepared to adapt their AI strategies accordingly. The controversy also emphasizes the importance of transparency in AI systems, suggesting that companies should prioritize explainable AI and clear communication about the capabilities and limitations of their AI tools.
This development underscores the growing tension between AI innovation and governance. As language models become increasingly powerful and capable of generating realistic images, the potential for misuse and harm grows exponentially. The Grok 2 release demonstrates the urgent need for industry-wide standards and potentially stronger regulatory frameworks to govern AI development and deployment.
The release also exposes the limitations of current content moderation strategies on social media platforms. X.com’s hands-off approach to moderation is being put to the test as AI-generated content becomes increasingly sophisticated and difficult to distinguish from human-created material. This challenge is likely to become more acute as AI technologies continue to advance.
As the situation unfolds, it’s clear that the release of Grok 2 marks a pivotal moment in the ongoing debate over AI governance and ethics. It highlights the dichotomy between Musk’s vision of unfettered AI development and the more cautious approach favored by much of the tech industry and AI research community.
The coming weeks will likely see increased calls for regulation and industry-wide standards for AI development. How xAI and other companies respond to this challenge could shape the future of AI governance. Policymakers may feel compelled to act, potentially accelerating the development of AI-specific regulations in the United States and other countries.
For now, X.com users are grappling with a flood of AI-generated content that pushes the boundaries of acceptability. The incident serves as a stark reminder of the power of these technologies and the responsibility that comes with their development and deployment. As AI continues to advance rapidly, the tech industry, policymakers, and society at large must confront the complex challenges of ensuring these powerful tools are used responsibly and ethically.
Haters will say this is AI 🕺🕺 pic.twitter.com/vqWVxiYXeD
— Elon Musk (@elonmusk) August 14, 2024