AI has hit human-level performance on some parameters: Stanford report
Briefly

In 2023, 149 new foundation models were released, with 65.7% open-source compared to 44.4% in 2022 and closed-source models outperforming open-source by 24.2%.
Google led in AI with 40 foundational models; Gemini Ultra was the costliest at $191 million, highlighting significant investment in the field.
The US excelled in AI development in 2023, creating 61 models and leading in investment with $67.2 billion, surpassing China by 8.7 times.
AI has achieved human-level performance in various benchmarks like testing, reading, comprehension, and visual reasoning, showcasing advancements in the field.
Read at The Economic Times
[
add
]
[
|
|
]