161–170 of 283 posts
Article from a slightly better source: https://timesofindia.indiatimes.com/technology/tech-news/how...
From what I read online, the real issue was "Natasha", their virtual assistant powered by a dedicated foundation model. They ran out of money before it got anywhere.
Earlier quoted context omitted.
$400M! I get $100M. Maybe even $200M. But $400M? Unforgivable.
You figure 700 employees. 400m. Avg cost per hooker can't be more than a few hundred. So by this math each employee got 1,900ish hookers. Since i figure male hookers for the female employees where cheaper well round up to 2,000. That is in fact unforgivable. 1,000 would of been acceptable. 2,000... just excess
Shameful.
AI = Actual Indians, apparently.
Yep. Also happened with Amazon's "Just Walk Out Technology:" https://www.businessinsider.com/amazons-just-walk-out-actual...
I worked on the first iteration of Amazon Go in 2015/16 and can provide some context on the human oversight aspects.
The system incorporated human review in two primary capacities:
1. Low-confidence event resolution: A subset of customer interactions resulted in low-confidence classifications that were routed to human reviewers for verification. These events typically involved edge cases that were challenging for the automated systems to resolve definitively. The proportion of these events was expected to decrease over time as the models improved. This was my experience during my time with Go.
2. Training data generation: Human annotators played a significant role in labeling interactions for model training-- particularly when introducing new store fixtures or customer behaviors. For instance, when new equipment like coffee machines were added, the system would initially flag all related interactions for human annotation to build training datasets for those specific use cases. Of course, that results in a surge of humans needed for annotation while the data is collected.
Scaling from smaller grab-and-go formats to larger retail environments (Fresh, Whole Foods) would require expanded annotation efforts due to the increased complexity and variety of customer interactions in those settings.
This approach represents a fairly standard machine learning deployment pattern where human oversight serves both quality assurance and continuous improvement.
The news story is entertaining but it implies there was no working tech behind Amazon Go which just isn't true.
Article from a slightly better source: https://timesofindia.indiatimes.com/technology/tech-news/how...
This is fake news. Builder.ai, like any other dev shop, had clients and was building apps using developers in India, pretty much like Infosys or any other Indian dev shop. Nothing wrong with that. From what I read online, the real issue was "Natasha", their virtual assistant powered by a dedicated foundation model. They ran out of money before it got anywhere.
Yeeeah... that's a fairly disingenuous take.
The difference between every other offshore dev shop backed by developers in India and Builder.ai is that - and I say this as someone who thinks Infosys is a shit company - Infosys and all those other dev shops are at least up front about how their business works and where and who will be building your app. Whereas Builder.ai spent quite a long time pretending like they had AI doing the work when actually it was a lot of devs in India.
That is deliberately misleading and it is not OK. It's fraudulent. It's literally what Theranos did with their Edison machines that never worked so whereas they claimed they had this wondrous new blood testing technology they were actually running tests with Siemens machines, diluting blood samples, etc. The consequences of Theranos's actions were much more serious (misdiagnoses and, indeed missed diagnoses of thousands of patients), rather than just apps built by humans rather than AI, but lying and fraud is lying and fraud.
Article from a slightly better source: https://timesofindia.indiatimes.com/technology/tech-news/how...
So.. where did the $450M go? A team of 700 developers in India built over eight years would have cost a fraction of that.
Earlier quoted context omitted.
DeepSeek cost just over $5M to train. StarCoder cost around $1M, there is no info for Starcoder2 but unlikely to be more than a few million. The idea of spending billions in training is OpenAI trying to build a moat that might not actually exist.
These architectures didn't exist last year. The Chinese are innovating thanks to massive government backing, access to talent, and a clear focus on winning the AI race.
That was right around the time this company had a new $250M funding round, so lack of resources to invest in actual AI is a terrible excuse.