SpaceX Sabotaged Ukraine’s Drone Attacks, The Shane Gillis Situation is INSANE, The FBI Arrested an OpenAI Engineer | Weekly Recap

Ukraine’s defense-technology adviser confirmed that speed restrictions had been applied to military Starlink terminals, amid reports that drone missions failed or were cancelled after connectivity interruptions of up to 20 seconds. Reuters reported disruptions at speeds as low as 68 mph, while the reported rule cuts connections above 100 mph. SpaceX and Starlink had not commented, and neither the actual limit nor the cause was clear.

Ukraine had initially requested a speed cap after Russian forces began equipping BM-35 strike drones and Shaheds with Starlink terminals to extend their range. Ukrainian officials called the disruption suspicious, but the discussion found no evidence it was politically motivated or deliberate. A technical glitch or a failed Ukrainian carve-out were also possible explanations. The timing overlapped with Trump’s opposition to Zelensky, but that alone does not establish a connection. Reports that Ukraine struck back after Russian attacks on energy infrastructure were unverified, and the outages may have prevented some Ukrainian attacks.

Shane Gillis’s joke about women dressing like “whores” at the gym prompted a wider argument about comedy, gender and responsibility. Levy called the joke misogynistic and said men’s complacency helps create an environment in which women are sexually assaulted. Another speaker questioned the jump from a joke to sexual assault, said it was not among Gillis’s most outrageous material, and argued that the SNL joke was lazy. In that view, its timing amid speculation about the Cornell seven mattered more than its delivery or content.

The speakers also disagreed over whether women face systemic barriers in comedy. One cited skipped showcases and hostile comments; another said anecdotes and online abuse do not establish systemic suppression or prevent women from finding audiences online. That speaker knew of no credible evidence that algorithms suppress women comedians or that women have more difficulty appealing to men through comedy. The discussion distinguished laughter from “clapter” and argued that people should examine why a joke resonates rather than berate others for finding it funny.

A parent described initially siding with a teacher who dress-coded Lynn for short shorts before cheerleading practice, then reconsidering after Lynn asked why she was responsible for boys’ reactions. The speakers agreed that schools need dress codes but differed on where to draw the line. They said boys should be taught not to stare, ogle, point or laugh, and that clothing does not excuse disrespect. One speaker said girls should understand the male gaze and consider the setting when choosing what to wear, while also saying women need not hide their attractiveness. The discussion held that provocative dress can draw attention and increase the risk of inappropriate behavior, but that such behavior remains wrong; it also noted that men are more likely than women to perpetrate sexual violence.

AI-safety whistleblower Jacob Coxon claimed China had spies inside OpenAI and Anthropic, estimating the likelihood at 90%. The host said espionage is common and AI is a strategic national-security and economic priority, but favored practical security measures developed by industry and government over relying mainly on regulation, which could entrench incumbent firms. The host also cautioned against panic, severing every relationship or discriminating against Chinese people. Speakers said an unreleased OpenAI frontier model may have solved a longstanding mathematics problem, and that AI can now decompile binary code, exposing software that was previously harder to extract.

The discussion questioned whether a $700 billion investment bet was worthwhile, calling AI a bubble by traditional definitions while stressing that its scale and timing were unknowable. The suggested approach was to retain some investment exposure and diversify rather than hold only cash. A Wall Street Journal article about an unnamed former OpenAI employee’s departure appeared 15–18 minutes before he announced it; one speaker saw the timing as suggestive of coordination, but not proof. They said the employee could sincerely believe the industry has serious problems, while noting that visibility and vested OpenAI shares might shape his incentives consciously or otherwise. The effect of any resulting regulation or industry decline on those shares remained uncertain.