I could imagine running multiple different AIs seeing if they agree if something is true.
I imagine this is similar to how Twitter's/X's community notes work. Something along the lines of when you have two accounts that disagree traditionally and they agree on something. That's how you know it's likely to be true.
So from that approach, you actually want to have two AIs that are on the other side of the spectrum of whatever you're trying to find out if it's true e.g. if a conservative AI and a liberal AI both agree that something is false, it's highly likely to be false.
This is extremely problematic. Much of news is highly syndicated, so what looks like 10 credible sources are actually just 2. I think the best you can aim for is empirical sources: receipts, videos, photo evidence, public disclosures directly at the source. Archiving is ok if disclosed.
I don't think LLMs can discern truth when the majority of news sources are incentivized primarily for views.
Empirical sources - would be a first candidate to be implemented in a skill.
And you are right about 10 credible sources are actually 2. It will look like 10 independent sources to it, real gap. Skill does prefer primary source over secondary. For example it'll prefer NASA article over blog post who cited NASA article.
I'll work on empirical sources improvement in a next releases.
You may have overlooked this part of the parent's comment which addresses your syndication issue:
> Something along the lines of when you have two accounts that disagree traditionally and they agree on something. That's how you know it's likely to be true.
Outfits that blindly parrot talking points would not be at odds and thus not be good candidates.
It also doesn't help that many news sources will knowingly fudge the truth. There are topics about which they are constant sources of misinformation. Seems like depending on the topic (anything science related) you wouldn't want to use news sources at all. Other topics, you probably only want to use news sources. And arranging all of this weighting of sources is going to be difficult and controversial as well. Doesn't mean it can't be done. Doesn't mean it wouldn't be valuable. But there is a lot to it and most of it isn't so much about technology or math as it is about understanding who should and shouldn't be considered authoritative about what topics.
I think this is the first real step from Stripe into token economics. They tried with a usage-based product, but from talking to stripe engineers, it sounds like there was less appetite than they thought. Could have been for many reasons, but I wonder if owning the actual request layer is a kind of missing ingredient?
If so many applications are just going to be token resellers (which, to be honest, a lot of very successful companies today are), one of the biggest issues is: if I'm charging my users $10 a month for unlimited AI usage, what does that really mean and look like? Do the other users subsidise the super users' use?
Big companies can stomach this uncertainty if they've raised a tonne of capital, but if you're a little solo-preneur or small startup/business (which Stripe historically favour), you have to know exactly how much this is going to cost you. Else, if you're off by a few zeros, this will bankrupt your company.
I had 2 MacBook Pros. One 2024 and one 2019. The 2024 one would connect fine to the internet, the 2019 one would not.
After pasting in the airportd logs of both (into ChatGPT and Gemini) it found it was down to band switching (2.4GHz and 5GHz) through some really old error code.
This fixed a problem that had plagued me for >12 months. Really magical feeling it got in on first try.
One trick I've found is that I need to first give the clanker the real problem (in your case, the laptops not connecting) and then give it what I think is wrong. If I start at too low a level, I lose much of the advantage of the tool, which is part rubber ducky, part "did you think to try".
No, "grass always looks greener on the other side" is a perspective thing. If you stand on your own grass then you look down onto it and see the dirt, but if you look over to the other side you see the gras from the side which makes it look more dense and hides the dirt. But it's the same boring grass everywhere. :)
At first, I thought "this is missing the point of the phrase" and moved on, but now I'm back to say it's stuck in my head and an intuitive, pretty neat way to think about it.
I think there's some goldilocks speed limit for using these tools relative to your skillset. When you're building, you forget that you're also learning - which is why I actually favour some AI code editors that aren't as powerful because it gets me to stop and think.
I think this is the wrong way to think about it. In this case the "intelligent people who are wasted on finance and ads" are drawn to high-status, low-risk, well-paid jobs, with interesting problems to solve.
If you want to solve meaningful problems you need a different kind of intelligence; you need to be open to risk, have a lot of naivety, not status orientated, and a rare ability to see the forest among the trees (i.e. an interesting problem isn't necessarily a important one).
While true, another is that “crop harvesting efficiency” and medicine are both more a biology/chemistry problem which may not interest the same people so it’s unclear they’d even attract the same thing.
It’s also missing that advancements in one field, particularly computer science, computation, and AI creates significant infrastructure that can be applied to those tasks in never before seen ways.
And finally, physical problems evolve much more slowly and is more capital intensive and requires a lot more convincing of other people. Digital problems by comparison are more “shut up I’m right, here’s the code that does X”. It’s easier to validate, easier feedback resulting in quicker mastery, etc. Not saying it’s completely bulletproof in that way, but more true than in physical sciences these days. So just throwing more people at the problem may not necessarily yield results without correct funding which historically was provided by the government (hence the huge boom in the 60s) but as the low hanging fruit were picked and government became more dysfunctional, this slowed to a crawl.
For example, personally I probably could have ended up working on fusion research if I had more economic security growing up and it felt like the nuclear industry was booming instead of constantly underdeveloped (both fission and fusion). But instead I’ve worked with computers because I felt like it was a boom segment of the economy (and it has largely been while I’ve worked) and the problems felt interesting (I’ve worked on embedded OSes, mobile OSes, ML, large distributed systems, databases, and now AI) and like there’s always interesting products to build to help improve the world.
Should we view those who chase status as a bad thing, or look to those who assign status that is then chased? If the average person cares more about who won last night's big game than some work done to improve medication, should we really have anything to say about those who decide to optimize their lives by what society actually rewards?
I noticed this way back in grade school. Good grades were, if anything, a net negative prestige, while sports were a positive prestige. It made me wonder what the school was actually optimizing for, because the day to day rewards weren't being given to the studious. (The actual reward function was more complicated, such as good grades being a boost if one was already a sports star, but these were exceptions to the norm.)
No, that's just an optimization that saved on computing resources. It effectively allows the party that runs this simulation to have a limited world to simulate. Dark matter is the other half of that trick. Both were invented by one Bebele Zropaxhodb after a particularly interesting party in the universe just above this one...
I imagine this is similar to how Twitter's/X's community notes work. Something along the lines of when you have two accounts that disagree traditionally and they agree on something. That's how you know it's likely to be true.
So from that approach, you actually want to have two AIs that are on the other side of the spectrum of whatever you're trying to find out if it's true e.g. if a conservative AI and a liberal AI both agree that something is false, it's highly likely to be false.