Since people like an unrelated product, you thought it would be good marketing to jack the brand? Maybe I should start calling my B2B SaaS products Salesforce too!
Aviation is an incredibly low-margin business - airlines are less trying to earn "massive profits" and more just trying to continue to operate year over year
A household might have three or four mobile phones, but only one or two cars. Many cars also can and do support both Android Auto and Carplay. If my wife and I used Android and my teenage son used Apple, I'd definitely want our car to support both.
Say you get your way, and, for fear of Mark Carney rolling the tanks in and taking over North America, the US stops collecting any data on its citizens. How is the IRS supposed to know how much tax to expect from you? How is SNAP supposed to determine your eligibility? How is unemployment supposed to know if you're ripping them off or not? Data privacy is a real concern, but you need PII to run government services effectively. Running a state without collecting PII is like running a hospital without collecting any.
Your rhetoric is clever, but I see a hole in it. Indefinite detainment is a logical consequence of lacking a right to a speedy trial, but is not identical to it.
I think it is in fact quite apt to say that the right to a speedy trial is to indefinite detainment as the right to privacy is to your information being used against you. Both are material consequences of lacking those rights.
Could you expand more on your struggle session idea? I'm aware of the analogy to Mao's China but I'm curious how you'd see it playing out in this setting.
Perhaps I'm being overly contrarian, but from my point of view, I feel that could be a blessing in disguise. For example, in a world where deepfake pornography is ubiquitous, it becomes much harder to tarnish someone's reputation through revenge porn, real or fake. I'm reminded of Syndrome from The Incredibles: "When everyone is super no one will be."
I'm curious about how the economics of human feedback will work out in the long term. I currently work in operations at a data annotation company, and my experience has given me a very pessimistic view of the industry's current state.
The MO for these vendors and the AI companies that buy their data seems to be a race to the bottom in price with little concern for quality. The current industry norm is to outsource the work to developing countries, where the cost of living is more in line with what the annotation agencies are willing to pay. While this isn't necessarily problematic for quality in and of itself, it does seem to make it harder to find candidates with the English skills required to generate high-quality RLHF and SFT training data. Furthermore, the pay offered for coding annotators is not competitive with local pay for software engineers, making it challenging to recruit skilled programmers. A lot of coding annotation is done by beginners and students.
There is certainly a lot of hype surrounding LLMs and their potential to disrupt various industries — even the US DOD has been scoping out the potential use of LLMs to assist military commanders in strategic decisions. However, if we want these LLMs to consistently perform at an expert level, they need expert-level training data. I worry that producing this data at the quality and scale required may be prohibitively expensive, and could cause a major bottleneck in model improvements long-term.