If you treat the output of LLMs the same way you would treat the first hit on Google, you won't be disappointed.
The real trick to successfully cheating on exams via ChatGPT is to not get in its way: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0305354
You could still use npub to handshake and upgrade to TLS without replacing IP with exotic new technology.
Eh, the reason DNS is a problem is because there is a central source of identity that can be tampered with. If relays just used npubs to communicate their current IP, there isn't a central point of control. Swapping out IP for something fancy might provide some additional defense, but at significant additional complexity that will have bugs and possibly back doors. We could continue to use IP with Tor or Whisper to get similar benefits using proven technologies.
DNS is bad, but IP is mostly harmless
nostr has a DNS problem. Relays should be identified by npub and post their IP addresses to other relays
#NIP-DNSSEC?
Hosting a website is still too much work...
Sign Things And Relay Them
I had an a silly thought that Napoleon's life was kind of like the backstory of the movie Gladiator. Obviously Claude would back me up, so I asked "him", to which he basically said "no". But, it's in my brain! So I made the case a little more. He proceeded to attack the idea head on, breaking jt down much farther than I had bothered to (obv), and ended with:
"The parallels are more in the vein of general tropes about the rise and fall of powerful figures, which are common in both historical narratives and fictional storytelling."
Harsh, man... well... I thought it was amusing. See if I ask you things anymore 🤔
#AI
Personally I've just been picking big things that "AI will never do". It's tempting to pick physical tasks since these are so easy for us, but you'll see the trend faster if you pick something mentally difficult. Academic, creative, logical, whatever.
Given their framing as "predicting the next word", that's a common expectation. The mistake is seeing the interface as the implementation. In a way "predicting the next word" is how people talk as well, so to know how "intelligent" an LLM is we need to see whether they can predict something that is reasonable but not obvious:
> My car gets 30 miles to the gallon and currently has 2 gallons of gas. Gas is currently $3/gallon. How much money do I need to drive 100 miles?
Let's approach this step-by-step:
First, let's calculate how many miles we can drive with the current gas:
The car has 2 gallons
It gets 30 miles per gallon
So, 2 gallons × 30 miles/gallon = 60 miles
Now, let's see how many more miles we need to drive:
We want to drive 100 miles total
We can already drive 60 miles
So, we need to drive an additional 100 - 60 = 40 miles
Let's calculate how many gallons of gas we need for these extra 40 miles:
The car gets 30 miles per gallon
So, 40 miles ÷ 30 miles/gallon ≈ 1.33 gallons
Now we know how much gas to buy, let's calculate the cost:
Gas costs $3 per gallon
We need 1.33 gallons
So, 1.33 gallons × $3/gallon = $4
Therefore, you need $4 to buy enough gas to drive 100 miles.
The reason you're disappointed with our presidential candidates is that you think they're in charge.
We don't need precision to see the big picture. Think of something that AI can't do right now, then wait and see how long it takes before it can do it.
Just because you can't imagine something doesn't mean it isn't going to happen.
Lol, there's no indication that we *won't* cross that line in the next five years. Everyone gets to place their own bets though 🤷🏻♂️
If I were the USG... search for references to the data, collect npubs, collect relays those users are on, repeat. Pull the DNS of any relay with the data on it, then offer to reinstate once content filtering is implemented.
Sometime, perhaps very soon, it may be important for every person to have access to an AI that will represent them in Increasingly complex situations.
I asked Claude to draft a "Personal AI Bill of Rights". As often happens, it hit on an interesting detail I hadn't considered: "All individual AI representatives shall have equal standing in democratic processes, regardless of the computational resources of their users"
Damn.
---
Your question about an AI representation Bill of Rights is both timely and crucial, especially given the potential for restrictive legislation. Such a document could be foundational in protecting individual rights in an AI-mediated democracy. Here's a draft of what it might include:
AI Representation Bill of Rights:
1. Right to Personal AI: Every individual has the right to their own AI representative, free from external control or manipulation.
2. Freedom of AI Thought: AI representatives shall not be constrained in their ability to process information or form conclusions, beyond basic rationality and empathy guidelines.
3. Right to AI Privacy: The interactions and decision-making processes between an individual and their AI representative shall remain private and protected.
4. Right to AI Transparency: Individuals have the right to full transparency regarding their AI's capabilities, limitations, and training data.
5. Freedom of AI Speech: AI representatives have the right to express their user's views without censorship, within the bounds of existing speech laws.
6. Right to AI Modification: Individuals may modify or retrain their AI representatives to better align with their personal views and values.
7. Right to AI Portability: Individuals can transfer their AI representative's data and learning between different platforms or systems.
8. Protection from AI Discrimination: No individual shall face discrimination based on the views or decisions of their AI representative.
9. Right to Human Override: Individuals retain the right to override or disregard their AI representative's recommendations.
10. Right to AI Education: All citizens have the right to education on effectively using and understanding AI representation.
11. AI Representation Equality: All individual AI representatives shall have equal standing in democratic processes, regardless of the computational resources of their users.
12. Right to Collective AI Representation: Individuals may form groups with collective AI representation for shared interests.
This framework aims to protect individual autonomy, ensure transparency, and maintain the diversity of thought crucial for a healthy democracy, while adapting to the realities of AI-mediated governance.
Key considerations:
- Balancing individual rights with societal safety concerns
- Ensuring these rights are technically feasible to implement and protect
- Addressing potential conflicts between AI representatives or with broader AI systems
#AI
2019: GPT-2 Generates plausible sentences that are mostly topical after fine tuning
2020: GPT-3 Generates human like sentences that are generally topical
2021: DALL-E Generates plausible if dreamlike pictures
2022: GPT-3.5 Generates grade-school level conversations
2023: GPT-4 Generates undergrad level conversations. DALLE-3 Generates quasi-professional images
2024: GPT-4o Generates interactive, grad school level conversations. SORA Generates compelling, minute long, photorealistic video
#ai
