"In a recent post on social media platform X, Greg Kamradt, President of the ARC Prize Foundation, noted that Grok 4 has been used in two SOTA submissions to the ARC-AGI V1 and V2 benchmarks. The submissions, which were the best that the foundation has encountered to date, were done using xAI's large language model, Grok 4."
"Elon Musk has shared a number of tidbits about Grok 5 as of late. In recent comments on X, the Tesla and SpaceX CEO stated that he believes Grok 5 has the potential to actually reach artificial general intelligence (AGI). This suggests that Grok 5 would be nothing short of a monster, as it could be capable of matching or surpassing human-level intelligence across tasks such as thinking, reasoning, and other domains by a substantial margin. Grok 5 may also be coming soon. As per Musk in another post on X, xAI will be starting the training of Grok 5 in a few weeks. Musk definitely seems to be highly optimistic about the capabilities of Grok 5, with the CEO stating late last month that the update would be coming before the end of the year and that it will be "crushingly good.""
Grok 4 has been used in two state-of-the-art submissions to the ARC-AGI V1 and V2 benchmarks and produced the best results encountered by the foundation to date. The authors of those submissions reported that Grok 4 was the best model used in testing. xAI plans to begin training Grok 5 in the near future, with training expected to start in a few weeks and an update anticipated before year-end. Grok 5 is described as potentially capable of reaching artificial general intelligence and is characterized as likely to be exceptionally powerful.
Read at TESLARATI
Unable to calculate read time
Collection
[
|
...
]