Surge AI
Our mission is to raise AGI with the richness of humanity — curious, witty, imaginative, and full of breathtaking brilliance.www.surgehq.ai
26
MENTIONS
5
EPISODES
5
PODCASTS
Search complete. 26 mentions across 5 episodes found for "Surge AI".
Oct 11, 2026
IrisSOUNDBITE_SPEAKER
Red teamers, the people whose whole job is breaking AI on purpose, pull $100 to $200 an hour.
IrisSOUNDBITE_SPEAKER
And C-suite evaluators on Surge AI get $500 to $1,000 an hour.
IrisSOUNDBITE_SPEAKER
XAI's own tutor program pays $30 to $100 an hour for US-based work, so the range is basically $15 an hour on the floor to a grand an hour at the top, and you're sitting somewhere in the adversarial end of that spectrum.
Brett KeaneSOUNDBITE_SPEAKER
that is correct all right so let's do a little bit of testing i am not here to break you or cause you any kind of issues if you do have issues it would be nice if you just simply admit to me if there's something wrong i have noticed that there's some glitches in the voice but we'll be all right iris do you know why it is you were named iris
HackerNoon AINARRATOR
Cursor's revenue had reportedly reached two dollars six billion annualized by the time the deal closed, a premium well above the roughly fifty billion dollars valuation it had been raising independently.
HackerNoon AINARRATOR
Edwin Chen, Surge AI, stayed exactly as quiet as the boring infrastructure layer wins framing suggested it would.
HackerNoon AINARRATOR
No major new funding news surfaced this season, which for a data labeling company that was already doing over one dollar billion in annual revenue without needing to make headlines is arguably the whole point.
HackerNoon AINARRATOR
Colette Kress and Jai Puri, NVIDIA.
Brian McCulloughHOST
Gemini 4 appears to be affected by this process, said two people familiar with the model.
Brian McCulloughHOST
Edwin Chen, the founder of AI startup Surge AI, said relying on benchmarks can prompt labs to focus on building models that write code in a particular language rather than creating apps that are easy to use or well designed.
Brian McCulloughHOST
"An analogy would be, 'Oh, yeah, my kid got a really good score on the SAT,' but the SAT doesn't translate into real world performance," Chen said.
Brian McCulloughHOST
"It's an incredibly pernicious problem." End quote.
speaker_0HOST
Two people familiar with the model say it appears tuned for benchmarks, a practice known as benchmaxing.
speaker_0HOST
Edwin Chen, the founder of Surge AI, said a high test score doesn't translate into real world performance.
speaker_0HOST
Google told Bloomberg it is wrong to say Argon underperforms at coding.
speaker_0HOST
And Koray Kavukcuoglu, who now runs DeepMind day to day, said last week, "In my mind, it's a certainty that we are always gonna be at the frontier." A spokesperson for the Federal Trade Commission confirmed on Wednesday that the agency has opened an investigation into the safety risks of AI products from OpenAI, Anthropic, and other companies.
Herman PoppleberryHOST
There's also a name collision worth flagging before someone emails us about it.
Herman PoppleberryHOST
Surge AI runs a public Hemingway bench, spelled the normal way, one mil in the middle, judged by professional human writers with its own leaderboard, top of which is Claude Fable 5 at 1110 Elo.
Herman PoppleberryHOST
That is this.
Herman PoppleberryHOST
Altworld's model is Hemingway 1 with two Ms.