Resolves yes if xAI is commonly perceived as being in the same tier as OpenAI, DeepMind, and Anthropic on January 1, 2025.
Inherently subjective, but factors to be considered are: funding, technical breakthroughs, new foundation models, mainstream products, mentions alongside other top labs in media, and its CEO's presence alongside other major lab CEOs at major meetings with e.g. world leaders, plus anything else that seems pertinent.
Resolves no if the criteria are never met, or if they're met, but xAI goes bust or gets absorbed by a pre-existing top lab before the close date. If xAI gets bought by a then-not-major lab/player (e.g. Apple), the market resolves yes if the criteria have otherwise been met.
I won't trade in this market.
1,000🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ6,277 | |
| 2 | Ṁ3,235 | |
| 3 | Ṁ2,580 | |
| 4 | Ṁ1,886 | |
| 5 | Ṁ1,374 |
funding $50 billion valuation. $8+B,
technical breakthroughs (ethernet and other advances with for AI data centers)
new foundation models (various 2nd and 1st place rankings throughout 2024)
mainstream products (Grok in xAI used by hundreds of millions
mentions alongside other top labs in media (Sam Altman says xAI is a serious competitor, Nvidia Jensen Huang is amazed at their AI data center and installation)
CEO's presence alongside other major lab CEOs at major meetings with e.g. world leaders, plus anything else that seems pertinent. (Elon sits with ALL World leaders and talks AI with them. Sacks is an xAI shareholder and he is the AI Czar). Sacks is key to setting policy for US and the world.
@brianwang Perplexity says YES.
xAI was indeed commonly perceived as being in the same tier as OpenAI, DeepMind, and Anthropic by January 1, 2025. Here's an analysis of the key factors:
1. Funding and Valuation: xAI raised significant capital, with a $5 billion funding round in November 2024 valuing the company at $50 billion[1]. This was followed by another $6 billion raise in December 2024[4][8]. Such substantial funding puts xAI in the same league as other top AI labs financially.
2. Technical Breakthroughs: xAI made notable advancements in AI infrastructure, particularly with the rapid deployment of their Colossus supercomputer cluster and innovations in scaling GPU clusters for AI training[1].
3. New Foundation Models: Throughout 2024, xAI's Grok models showed competitive performance on various benchmarks. Grok-1.5V, released in April 2024, demonstrated leading performance in multimodal tasks, outperforming GPT-4V, Claude 3, and Gemini Pro 1.5 in certain areas[3].
4. Mainstream Products: Grok AI was made available to X Premium+ subscribers, potentially reaching a large user base[2].
5. Media Mentions: xAI has been frequently mentioned alongside other top AI labs in media coverage, with comparisons often drawn between Grok and ChatGPT[2][5].
6. CEO Presence: While not explicitly mentioned in the search results, Elon Musk's high profile and connections to world leaders likely contributed to xAI's perceived status.
7. Industry Recognition: Nvidia CEO Jensen Huang's praise for xAI's rapid deployment of the Colossus cluster indicates recognition from industry leaders[1].
8. Competitive Performance: Grok models have shown competitive performance against established models from OpenAI and Anthropic in various benchmarks[2][3][5].
9. Rapid Growth: xAI's quick ascent to a high valuation and its ability to attract significant investment demonstrate its perceived potential in the AI industry[1][4][8].
While the search results don't provide explicit information about xAI's status on January 1, 2025, the developments throughout 2024 suggest that xAI had established itself as a serious competitor in the AI space. The combination of substantial funding, technical achievements, competitive models, and industry recognition indicates that xAI was likely perceived as being in the same tier as other leading AI labs by the beginning of 2025.
Citations:
[1] https://www.pymnts.com/news/2024/report-elon-musks-xai-raises-5-billion-and-doubles-valuation/
[2] https://em360tech.com/tech-articles/xais-grok-better-chatgpt-comparison
[3] https://encord.com/blog/elon-musk-xai-grok-15-vision/
[4] https://www.nytimes.com/2024/12/24/technology/elon-musk-xai-funding.html
[5] https://www.superannotate.com/blog/grok-ai-elon-musk
[7] https://synthedia.substack.com/p/xai-announces-grok-15v-multimodal
[10] https://www.datacenterdynamics.com/en/news/xai-raises-another-6bn-including-from-nvidia-and-amd/
@MattReardon In August 2024, Grok-2 secured second place on the LMSys chatbot Arena leaderboard, outperforming OpenAI's GPT-4o (May) and tying with the latest Gemini model
https://www.aibase.com/news/11268
xAI made innovations in speed of AI data center installation. Nvidia Jensen Huang said they were superhuman with the innovations they made to install it.
xAI rewrote two layers of the ethernet stack into hardware to enable faster communication to enable larger cohesive memory and AI chip mesh.
Those achievements were monumental research and implementation achievements.
Grok-2 excelled particularly in mathematical tasks, claiming the top spot in that category on the LMSys leaderboard
An early version of Grok-2, tested under the name "sus-column-r" on the LMSYS leaderboard, outperformed both Claude 3.5 Sonnet and GPT-4 Turbo
https://www.cmswire.com/digital-experience/what-is-grok-elon-musks-rebellious-new-ai/
In January 2025, Grok-2 was reported to have exceeded the performance of Claude 3.5 Sonnet and GPT-4 Turbo on the MMLU (Massive Multitask Language Understanding) benchmark3.
https://www.grokmountain.com/p/grok-models-vs-the-giants-performance
Grok-2 also demonstrated superior performance on the BigBench Hard (BBH) benchmark, outperforming established models like Claude 3.5 Sonnet and GPT-4 Turbo in certain aspects
@MattReardon Challenges of scaling to 100,000 H100 GPUs in one coherent data center.
June 2024
https://semianalysis.com/2024/06/17/100000-h100-clusters-power-network/
Patel talking about making the 100k and now 200K xAI colossus coherent memory AI data center
https://www.youtube.com/watch?app=desktop&v=8w4tohBnJN4
Gavin Baker the xAI data center innovations
https://www.youtube.com/watch?v=K2xfW3hgxb4&t=2948s
Sam Altman says xAI is a serious competitor
https://www.youtube.com/watch?v=RcOEdHOZRkQ
Jensen on xAI and Elon
Patent was Tesla but xAI and Tesla are siamese twins for AI.
https://x.com/seti_park/status/1865948153138348482
https://x.com/seti_park/status/1813213756153008336
@MattReardon yeah could go either way but your resolution is reasonable.
In particular I think xAI is the 4-6th lab in significance and I think very few reasonable observers would say xAI was a tier above any of OpenAI, GDM, or Anthropic on January 1. So it's either the top tier with 4+ labs (possibly also DeepSeek) or the top tier is just the top three.
@JoshYou Very reasonable case that they are above Anthropic on Jan 1 based upon funding, product usage, policy influence, AI data center technology, media . Plus they were above Google and Anthropic in August. The question said along side other top models OpenAI, Google, Anthropic. This means top 4 in the same tier. 4th very close to any of the top 3.
Asked the LLMs to decode the phrasing.
Being in the same tier as OpenAI, DeepMind, and Anthropic means the company is regarded as an influential AI leader but may not yet have the same legacy or consistency as the top 3. A "top 4" tier signifies it is on the rise, but not necessarily recognized as the dominant force in the industry just yet. It implies proximity to the top 3 but not yet fully in the lead.
@AndrewMcKnight @MattReardon probably gets a little bit of an extension for writing a comment here, but it has been 22 days. If there is no progress soon, we might need to call for a panel of 3 moderators to resolve the question -- but I would rather have @MattReardon resolve it because the moderators are probably not experts in this topic.
@AndrewMcKnight longer we wait on this, the more current event vibes from January and beyond will seep into the resolution decision
Definitely not commonly perceived as being in the same tier as OpenAI DeepMind, or Anthropic. Their models are used all over the place. Where is Grok 2 used? If Grok 3 is a success then that will push them up there. But it only finished pre-training in 2025.
I think xAI is definitely a major AI lab. Has a bigger computer than almost anyone else. Has lots of funding. Although the integration into X is a wash IMO -- people don't really use Grok chat much AFAICT.
Anyway, despite it being a major lab, according to the resolution criteria I think it's a NO.
@DavidFWatson I don't see where Matt ever used the word "Meta." Surely not mentioning Meta is not the same as saying "Meta explicitly doesn’t qualify."
From the little clarification Matt has given, where reasonable traders came to different conclusions about what it would mean for a lab to be "major," and where repeated requests for clarification were ignored (see unanswered question from @AndrewMcKnight), I still think the best solution is neither a 0% or 100% resolution.
@AdamK I believe that by listing the qualifying entities as of when the market was creating, the author was implicitly saying that those are the only leading edge AI labs at that time.
@DavidFWatson "explicitly" (your first comment) is the antonym of "implicitly" (your second comment). These can't both be true, and explicit exclusion would be a much more compelling argument than implicit exclusion.
@Jacy You're absolutely right that it's a bit of a contradiction, but to better clarify what I meant:
Explicitly: The description doesn't just say "is xAI a major AI lab" it goes further by explicitly listing major AI labs.
Implicitly: While the list doesn't explicitly list every lab that doesn't qualify, it implies that Meta isn't one of them by not listing it.