1 · Evidence grading.
DeepSeek says its V4-Pro coding-agent work costs roughly 8 times less than OpenAI's Codex for similar output. What's the evidentiary status of that number, per the reporting?
A· It's a figure AlphaSignal independently verified with its own benchmark tests
B· It's a number OpenAI itself confirmed after testing DeepSeek's model
C· It's a peer-reviewed academic result published alongside the model
D· It's DeepSeek's own claim about its own product, not an independently verified benchmark — worth treating as a claim, not a settled fact
2 · Interested messenger.
Cerebras raised its annual targets citing strong AI chip demand, even as its shares fell 16%, per Reuters. Whose assessment is that "strong AI chip demand" claim?
A· Cerebras's own guidance about its own business outlook
B· An independent analyst report commissioned by Reuters
C· A joint statement from Cerebras and a chip-manufacturing partner
D· A government trade agency's economic forecast
3 · Catch the error.
Statements about Twitch's AI-training disclosure:
§1· Twitch confirmed that Amazon has been training its generative AI on streamers' broadcasts, VODs, clips and chat for years.
§2· Twitch's own product chief told TechCrunch that requiring opt-in consent would get zero takers.
§3· The opt-out Twitch introduced was already switched on as the default months before the policy was ever announced, automatically excluding every streamer's past broadcasts.
§4· The opt-out Twitch is offering now covers streams, VODs, clips and chat going forward.
§5· Twitch said streamers who choose not to opt out will receive a share of any AI licensing revenue generated from their content.