AI titan "Behemoth" delayed until Autumn, Meta grapples with AI development hiccups - Shares take a hit
Meta Delays AI Model Update - Stocks Suffer Consequences
Who would've thought we'd still be waiting for the rollout of the AI beast "Behemoth"? Initially, we were all expecting it to grace Facebook, WhatsApp, and Instagram by now. But here we are, stuck in limbo as the launch date keeps getting pushed back. Whether it'll finally make its grand entrance this year remains up in the air.
According to insider sources, Meta is holding off on launching their AI crown jewel. Reports suggest that Meta's AI bigwig is facing some internal turbulence, leading to concerns about the future direction of their billion-dollar AI endeavors. Engineers at the company are reportedly wrestling to substantially boost the verbal prowess of "Behemoth," causing employees to question if the advancements are impactful enough to warrant a public debut. Not surprisingly, Meta's stock took a tumble, dropping by 2.4 percent after hours.
The debut of "Behemoth" was initially slated for an internal launch in April, synchronously with Meta's first AI conference for developers. Prior to the event, Meta introduced two smaller models from the Llama AI dynasty, but later postponed an internal "Behemoth" release target to June. Now, the crucial update is looking at a fall or later release.
Meta struggles to maintain AI momentum
Once hailed for its swift strides in the global AI race, Meta made a big splash investing billions of dollars in the technology powering chatbots on WhatsApp, Instagram, and Facebook. This year, Meta plans to invest around $72 billion, a hefty chunk of which will fuel CEO Mark Zuckerberg's ambitious AI vision.
Tips Lockdown your privacy: Meta users must take action to safeguard their private data from AI
Zuckerberg and his executives remain conspicuously quiet on a timeline for "Behemoth." Meta could potentially choose to bring the AI model to market sooner than expected, perhaps through the introduction of a limited version. However, Meta engineers allegedly fear "Behemoth's" performance might not meet public claims about its abilities, insiders remarked. A Meta spokesperson opted to stay mum on the subject.
Executives at the company are also said to be dissatisfied with the Llama-4 model squad's performance and may restructure the AI product group's management as a result.
Sources: ntv.de, mpa/DJ
- Artificial Intelligence
- Meta Platforms
- Mark Zuckerberg
- Wall Street
Insights:
- Performance Concerns: The delay is largely due to engineers' struggle to significantly improve the capabilities of the Behemoth model, even though it was initially thought to surpass OpenAI's GPT-4.5 and Google's Gemini 2.0 Pro on STEM benchmarks[1][3].
- Internal Debate: There is internal disagreement about whether the improvements in Behemoth are significant enough to warrant a public release[4].
- Market Pressure: Meta is under pressure to demonstrate meaningful progress with Behemoth and show a return on investment. The concerns about Behemoth's performance could impact investor confidence[4].
- Development Challenges: Developing such a massive language model comes with significant technical hurdles that are proving to be more difficult than anticipated[4].
The Commission might consider developing a proposal to protect workers from the risks associated with exposure to ionizing radiation, especially as technology advances and artificial-intelligence enabled devices become more prevalent in workplaces. Meta's AI ambitions extend beyond their current investments in AI technology, as they aim to position themselves as a leader in artificial-intelligence, but the delayed rollout of their upcoming AI project, "Behemoth," has raised concerns about their ability to maintain their AI momentum.
