@ajitmishra He has been in other cheatings like the Virat mentioned in his press conference.
He is a generational talent, an exceptional batsman but no respect for how he conducts himself on the field.
So let me get this right.
Your team did something terrible. And everyone knew what they were doing was terrible.
They got caught and now one of them is talking.
And Langer is justifying the bravery of the one who took the blame. That’s shite mate.
Its national Australian team, not your gang where you punish the rat and praise the guy who took the fall.
The nation deserves to know what happened.
@kikisknees China would probably have similar capabilities within a few months of the US doing this. China will then mass produce these and start dropping Temu orders in bulk and then the US will start panicking like the EV race.
Everyone’s obsessed with Jensen Huang getting honey-trapped.
Spare a thought for the spy.
She got fucked literally… then figuratively when the whole thing blew up and her cover was gone.
She is going to have a hard time getting another job.
Probably the first spy in history to lose her job to AI.
I believed the same but I’ve changed my opinion. Turns out quiet is just one part and it has to backed with substance. When some of those quiet ones opened their mouth, I realised they are just fos, they weren’t quiet coz of some grand thinking but they simply didn’t have something to say etc.
I’d frame it as Smart people who can stay quiet and deliver quality discussions/thoughts when they speak
The truth is people aren't multitasking to be hyper-efficient, they're high on the tech.
Right now it feels less like productivity and more like being a kid who just landed at Disneyland, parents are begging him to go to sleep, but his eyes are wide open.
Most heavy users aren't asking "Is this worth it?". They're just hooked on the adrenaline rush of running 10 agents at once.
I think most people multi-tasking with agents could benefit from slowing down
most of the stuff you're doing wont really matter- what's your P0? what would it look like if you gave it your full attention?
Let’s be real, Delta isn't scared of OpenAI.
They’re terrified of Kevin’s 1-person side project sending 4 billion rapid-fire requests because his async loop got stuck in a while loop.
Once frontier labs and these companies actually sit down and build real protocols (and agent gateways), they’ll be begging you to use them.
Early days, everyone just put their seatbelts on.
notice..
> Delta is blocking AI agents from booking flights
> JPMorgan Chase and Charles Schwab block agents from accessing accounts
> Yelp requires paid data licensing for agents
> Amazon has blocked Grok bot and other shopping agents.
the internet is starting to block AI agents that were supposed to do everything for us
Hot take:
While effective in the short term, this position is vulnerable, frontier labs will inevitably expand beyond core model development into tailored enterprise solutions.
I don’t buy the notion that frontier labs will just stick to core research and leave custom enterprise models to startups.
The second model progress starts plateauing, those same labs will look at the money on the table and say:
“Here’s our Supply Chain model. Take 2 of our FDEs. In a few weeks your entire operation is AI-native.”
Same for finance, legal, manufacturing, you name it.
They have the talent, the compute, the brand, and the hunger.
After the initial funding glucose rush fades, they won’t leave a single crumb.
They will eat this entire sector.
Over the last few months at @river_ai_inc we’ve developed what we believe is the best system in the world for building your own internal custom models, driven by 10 years of experience training top models at DeepMind, OpenAI and xAI.
In order to best support our customers, we’ve decided to do something radical: We also deploy our API on-prem on your compute cluster, and if you don’t have compute we will help you find it.
Again and again we’ve seen that custom models designed for your own internal workflows outperform top frontier models on performance, at a fraction of the cost (1 percent or less).
We’re just at the beginning of the custom model era, but its already clear that in the future more than 99% of LLM tokens will be generated on efficient custom models. The economics of it are just too powerful, and it’s far easier than people think once you have the right tools in place.
The reason why small models will dominate is due to the physics of deep learning: A general model that can do everything needs more capacity and training it on the whole distribution of all tasks is more challenging, so the model will be larger and more expensive, and wont perform perfectly on every task. A small model trained on a targeted domain usually outperforms it.
@logic@Zoho@svembu Bit jumpy, aren’t you ?
What makes you believe leak happened after Zoho switch ?
As per seller’s own admission data was taken from an "abandoned server” and contains 2022-24 data.
Switch to Zoho happened much later in 25.
Are you going to ask Satya Nadela now?
@ShubhangiYaps@AnshumanAI Probably coz you didn’t drink the glass on your right and it was wasted.
Seriously, ignore the trolls. It was for build for fun, not to compete with Gmaps
Nobody talks about the real reason most builders quit.
It’s not skill. It’s not money. It’s not even time.
It’s shipping something nobody uses and convincing yourself that means you’re not good enough.
You are good enough.
The app was just wrong.
Build the next one.
Chinese AI models might not be at the top yet, but they’re stopping US models from having the whole space to themselves.
In a way, they’re keeping the AI bubble in check.
I’m glad we have both on one side it prevents a small group from monopolising, and on the other it pushes current leaders to keep improving so we get better, cheaper models.