Ved

Chief of Staff @ CheerioAI. BITSian, Scaling systems, engineering GTM, and building AI tools from first principles.

[Ved @vdchdry]

@anishthite I got it from searching in X and checking the latest posts

[Ved @vdchdry]

@0xDevShah How you seen how fast these frontier lab’s revenue is climbing up, by end 2027, Anthropic will have more revenue than Meta

[Ved @vdchdry]

@dhsusj144284 @haider1 Yes, I used astra with plus plan. It finished my limit in 1 hours of usage. Whereas opus on pro feels sufficient

[Ved @vdchdry]

New one just dropped

[Ⓥea ⛓️ @Miamalista]

@Ferahleen

[Ved @vdchdry]

@leshka_eth So ?
They are at $62 billion run rate in july 2026. Growth counts, valuations are nothing more than perceptions

[Ved @vdchdry]

@BrianMRey Yeah same for google now what?

[Ved @vdchdry]

@_ak_111 I use opus as a teacher model for Gpt 6.1 sol

[Ved @vdchdry]

@_ak_111 Not even remotely close in my testing

[Ved @vdchdry]

@India_Policy There’s something known as quantity, not chinese and turkish slop drones and missiles

[Ved @vdchdry]

@socdemmi Check UAE

[Ved @vdchdry]

@akarlin You don’t know shit, India has 3rd largest compute in the world and is the world’s largest AI implementator

[Ved @vdchdry]

@thsottiaux Please add automode, I had to change the config for it.

Also when I am typing the prompt I should be anle to navigate with my trackpad

[Ved @vdchdry]

@0xPrajwal_ Now compare paid members

[Ved @vdchdry]

@blueemi99 You are an open ai shill, opus 5.5 is much better than 6.1 sol

[Ved @vdchdry]

@pprownets2023 Use jack one by one, can lift it

[Ved @vdchdry]

@AdamHoltererer What

[Ved @vdchdry]

Pretty crazy how the US doesn’t care about the fact that Osama Bin Laden was found in Pakistan and they still supply Pakistan with everything it wants.

[Ved @vdchdry]

In my testing Opus 5.5 is superior to 6.1 Sol for coding stuff

[Ved @vdchdry]

@shazcodes Better than making another AI SaaS

[Ved @vdchdry]

@theo I mean they can be nerfed by tuning parameters, but I don’t see any changes in the model’s outputs, they will always wary because these models are not deterministic. It’s same as saying to score exact same marks in every tesg