technology
S&P Proposes Rule Changes for MegaCap Inclusion
S&P Dow Jones Indices proposed alterations to accelerate MegaCap companies into the S&P 500. Meanwhile, AI expert Andrej Karpathy highlighted LLM advancements, and executives from Berkshire, Lazard, Chevron, and Reddit shared insights on quarterly performance and strategies. These developments occurred amid broader market updates on May 1, 2026.
KA
SA
JE
IN
CnbcAudio version
Tap play to generate a narrated version.
S&P Dow Jones Indices proposed rule changes for MegaCap companies to join the S&P 500, reducing the public company requirement to six months from 12 months. The proposal also waives the minimum public float requirement of 10% and the profitability requirement of positive net income. If approved, these changes would take effect before market open on June 8, 2026.
Post by @karpathy on X
Under the proposed rules, SpaceX could be added to the S&P 500 as soon as six months after its initial public offering. This inclusion could trigger hundreds of billions of dollars in forced buying by index funds, according to a post by @SawyerMerritt on X. Separately, Andrej Karpathy participated in a fireside chat at Sequoia Ascent 2026 approximately one week before May 1, 2026.
Karpathy stated that large language models enable new horizons beyond speeding up existing tasks like coding. He described 'menugen' as an app fully implemented with LLMs, taking an image input and producing an image output without classical code. sh scripts, where installation instructions are written in words for LLMs to interpret and execute adaptively.
He stated that LLM knowledge bases enable computation over unstructured data from arbitrary sources and formats, which was impossible with classical code. Karpathy explained jaggedness in LLMs as the ability to coherently refactor a 100,000-line code base while also suggesting mundane tasks like walking to a car wash.
He attributed LLM jaggedness to verifiability of domains and economics influencing training data distributions during reinforcement learning.
Karpathy discussed the agent-native economy, decomposing products into sensors, actuators, and logic across computing paradigms, and making information legible to LLMs. Lazard CEO Peter Orszag addressed Q1 2026 results, the acquisition of Campbell Lutyens, the Iran war, and tariff refunds in a CNBC video 27 minutes before May 1, 2026.
