I disagree with the “limitations” they ascribe to the Turing test - if anything, they’re implementation issues. For example:
For instance, any of the games played during the test are imitation games designed to test whether or not a machine can imitate a human. The evaluators make decisions solely based on the language or tone of messages they receive.
There’s absolutely no reason why the evaluators shouldn’t take the content of the messages into account, and use it to judge the reasoning ability of whoever they’re chatting with.
No, I want a communal, collaboratively managed platform to recommend things to me based on an open source algorithm whose behavior I can adjust the way I want. Alas, this just isn’t a thing.
Just amongst the available options, the closed algorithm optimized for engagement has so far been better at showing me interesting things than an unfiltered chronological feed.
I know it’s a feature, and I know people on Mastodon care about it. And because of that it’s not for me. That’s fine. My point was, exactly because Mastodon is not for everyone, there’s no need to be derisive of the people who “flock to yet another corporate social media honeypot.”
All that talk about “safety guardrails” is essentially a call against open source - when models are open, people can always remove them. That’s the price of freedom. And we have seen time and time again how the benefits outweigh that price.
I wouldn’t say there’s a place to start. Once you start using programs that are configured through config files, learn about those config files in particular. Eventually, you might find that you prefer editing config files even for programs that have GUI settings - then you dive in more.
Regardless, once your config files become complex enough that you can’t quickly rewrite them if necessary, start looking for a dotfiles manager, tracking them in git, backing them up, etc…
I don’t know why you would expect a pattern-recognition engine to generate pseudo-random seeds, but the reason OpenAI disliked the prompt is that it caused GPT to start repeating itself, and this might cause it to start printing training data verbatim.
I generally back up the whole ~/.mozilla, and if I restore it after reinstall, everything is as it was. I’ve not tried isolating only the profile, seems pointlessly complicated.
Platforms. I don’t believe that the people who create, or invest in, large internet platforms such as Facebook, YouTube, Uber, Booking, Upwork, etc, have a natural or moral right of ownership to said platform. They should certainly receive returns on their investment - but they shouldn’t have full operational control. Instead, as the platofrm grows, operational control should slowly transition to its users. eventually, they should have the final say on, in the case of YouTube. what content in acceptable, what procedures should be used to remove unacceptable content, how to appeal, etc.
Employment. One of the big issues I see is that employees are under someone’s direct control for 1/3 of each day, and have to do what their boss says. And while they technically consented to that relationship, I don’t see that consent as freely given, because for most people there isn’t a viable alternative. This could be done through more worker cooperative, or encouraging freelancing. Even for people who decide to remain in traditional employment, they should have more official control than they do now.
AI. It seems many people here hate AI, but AI does have the potential for large productivity gains. And while, in the past, productivity gains have note resulted in less work, but rather higher GDP, we could always force the issue. After all, people did it ~100 years ago, and the economy didn’t collapse because of that.
I haven’t had that issue. I’ve heard that disabling adblockers resolves it. But people have said that spoofing their user agent to chrome also magically resolves it…
Many imagine future AI as an autonomous agent. I don’t think anyone will release that. Instead, I expect to see a generative AI like GPT-4, however one that produces super-smart responses.
This will create a situation where the amount of computing resources someone has access to determines how much intelligence they can use. And the difference will be much bigger and more comprehensive than the difference between a genius and a normal human.