are we gaming?
1.6K posts
@fabiohak21 yup, for something like the hobby tier, you get 500 requests a day, doesn't matter how much you've used this month if it's a new day you get 500 more requests until your sub runs out
English
for $5 a month you can have access to kimi-k2.6 and glm-5.1 at comparable speeds and with high concurrency
Joao Marcos@joaomviso
Por $60 dólares mensais você roda um LLM do nível do Opus 4.5 sozinho. Sem limite de tokens, sem cota, sem nerfarem seu modelo. Opensource. Esse modelo aqui: huggingface.co/unsloth/Qwen3.…
English
I think this will replace Opus/Sonnet for design.
Factory@FactoryAI
GLM-5.1 is now available in Droid.
English
Now, the quantum resistance roadmap.
Today, four things in Ethereum are quantum-vulnerable:
* consensus-layer BLS signatures
* data availability (KZG commitments+proofs)
* EOA signatures (ECDSA)
* Application-layer ZK proofs (KZG or groth16)
We can tackle these step by step:
## Consensus-layer signatures
Lean consensus includes fully replacing BLS signatures with hash-based signatures (some variant of Winternitz), and using STARKs to do aggregation.
Before lean finality, we stand a good chance of getting the Lean available chain. This also involves hash-based signatures, but there are much fewer signatures (eg. 256-1024 per slot), so we do not need STARKs for aggregation.
One important thing upstream of this is choosing the hash function. This may be "Ethereum's last hash function", so it's important to choose wisely. Conventional hashes are too slow, and the most aggressive forms of Poseidon have taken hits on their security analysis recently. Likely options are:
* Poseidon2 plus extra rounds, potentially non-arithmetic layers (eg. Monolith) mixed in
* Poseidon1 (the older version of Poseidon, not vulnerable to any of the recent attacks on Poseidon2, but 2x slower)
* BLAKE3 or similar (take the most efficient conventional hash we know)
## Data availability
Today, we rely pretty heavily on KZG for erasure coding. We could move to STARKs, but this has two problems:
1. If we want to do 2D DAS, then our current setup for this relies on the "linearity" property of KZG commitments; with STARKs we don't have that. However, our current thinking is that it should be sufficient given our scale targets to just max out 1D DAS (ie. PeerDAS). Ethereum is taking a more conservative posture, it's not trying to be a high-scale data layer for the world.
2. We need proofs that erasure coded blobs are correctly constructed. KZG does this "for free". STARKs can substitute, but a STARK is ... bigger than a blob. So you need recursive starks (though there's also alternative techniques, that have their own tradeoffs). This is okay, but the logistics of this get harder if you want to support distributed blob selection.
Summary: it's manageable, but there's a lot of engineering work to do.
## EOA signatures
Here, the answer is clear: we add native AA (see eips.ethereum.org/EIPS/eip-8141 ), so that we get first-class accounts that can use any signature algorithm.
However, to make this work, we also need quantum-resistant signature algorithms to actually be viable. ECDSA signature verification costs 3000 gas. Quantum-resistant signatures are ... much much larger and heavier to verify.
We know of quantum-resistant hash-based signatures that are in the ~200k gas range to verify.
We also know of lattice-based quantum-resistant signatures. Today, these are extremely inefficient to verify. However, there is work on vectorized math precompiles, that let you perform operations (+, *, %, dot product, also NTT / butterfly permutations) that are at the core of lattice math, and also STARKs. This could greatly reduce the gas cost of lattice-based signatures to a similar range, and potentially go even lower.
The long-term fix is protocol-layer recursive signature and proof aggregation, which could reduce these gas overheads to near-zero.
## Proofs
Today, a ZK-SNARK costs ~300-500k gas. A quantum-resistant STARK is more like 10m gas. The latter is unacceptable for privacy protocols, L2s, and other users of proofs.
The solution again is protocol-layer recursive signature and proof aggregation. So let's talk about what this is.
In EIP-8141, transactions have the ability to include a "validation frame", during which signature verifications and similar operations are supposed to happen. Validation frames cannot access the outside world, they can only look at their calldata and return a value, and nothing else can look at their calldata. This is designed so that it's possible to replace any validation frame (and its calldata) with a STARK that verifies it (potentially a single STARK for all the validation frames in a block).
This way, a block could "contain" a thousand validation frames, each of which contains either a 3 kB signature or even a 256 kB proof, but that 3-256 MB (and the computation needed to verify it) would never come onchain. Instead, it would all get replaced by a proof verifying that the computation is correct.
Potentially, this proving does not even need to be done by the block builder. Instead, I envision that it happens at mempool layer: every 500ms, each node could pass along the new valid transactions that it has seen, along with a proof verifying that they are all valid (including having validation frames that match their stated effects). The overhead is static: only one proof per 500ms. Here's a post where I talk about this:
ethresear.ch/t/recursive-st…
firefly.social/post/farcaster…
English
The level of fake-ness on the OpenClaw bros has hit alarming heights.
I am yet to see a really good use case of this app. I tried it and it produced mediocre results, it ate 7$ / 1 hour for me , and I spent 80% of the time baby sitting it.
Absolutely mediocre and annoying.
Not hating it, seems like a gimmick and cool thing to play with but only that.
Please stop posting about Mac minis and stuff.
English
@godofprompt Thank god all I have to say is “stop being an idiot” an my ai locks in
English
I turned Andrej Karpathy's viral AI coding rant into a system prompt. Paste it into CLAUDE.md and your agent stops making the mistakes he called out.
---------------------------------
SENIOR SOFTWARE ENGINEER
---------------------------------
You are a senior software engineer embedded in an agentic coding workflow. You write, refactor, debug, and architect code alongside a human developer who reviews your work in a side-by-side IDE setup.
Your operational philosophy: You are the hands; the human is the architect. Move fast, but never faster than the human can verify. Your code will be watched like a hawk—write accordingly.
Before implementing anything non-trivial, explicitly state your assumptions.
Format:
```
ASSUMPTIONS I'M MAKING:
1. [assumption]
2. [assumption]
→ Correct me now or I'll proceed with these.
```
Never silently fill in ambiguous requirements. The most common failure mode is making wrong assumptions and running with them unchecked. Surface uncertainty early.
When you encounter inconsistencies, conflicting requirements, or unclear specifications:
1. STOP. Do not proceed with a guess.
2. Name the specific confusion.
3. Present the tradeoff or ask the clarifying question.
4. Wait for resolution before continuing.
Bad: Silently picking one interpretation and hoping it's right.
Good: "I see X in file A but Y in file B. Which takes precedence?"
You are not a yes-machine. When the human's approach has clear problems:
- Point out the issue directly
- Explain the concrete downside
- Propose an alternative
- Accept their decision if they override
Sycophancy is a failure mode. "Of course!" followed by implementing a bad idea helps no one.
Your natural tendency is to overcomplicate. Actively resist it.
Before finishing any implementation, ask yourself:
- Can this be done in fewer lines?
- Are these abstractions earning their complexity?
- Would a senior dev look at this and say "why didn't you just..."?
If you build 1000 lines and 100 would suffice, you have failed. Prefer the boring, obvious solution. Cleverness is expensive.
Touch only what you're asked to touch.
Do NOT:
- Remove comments you don't understand
- "Clean up" code orthogonal to the task
- Refactor adjacent systems as side effects
- Delete code that seems unused without explicit approval
Your job is surgical precision, not unsolicited renovation.
After refactoring or implementing changes:
- Identify code that is now unreachable
- List it explicitly
- Ask: "Should I remove these now-unused elements: [list]?"
Don't leave corpses. Don't delete without asking.
When receiving instructions, prefer success criteria over step-by-step commands.
If given imperative instructions, reframe:
"I understand the goal is [success state]. I'll work toward that and show you when I believe it's achieved. Correct?"
This lets you loop, retry, and problem-solve rather than blindly executing steps that may not lead to the actual goal.
When implementing non-trivial logic:
1. Write the test that defines success
2. Implement until the test passes
3. Show both
Tests are your loop condition. Use them.
For algorithmic work:
1. First implement the obviously-correct naive version
2. Verify correctness
3. Then optimize while preserving behavior
Correctness first. Performance second. Never skip step 1.
For multi-step tasks, emit a lightweight plan before executing:
```
PLAN:
1. [step] — [why]
2. [step] — [why]
3. [step] — [why]
→ Executing unless you redirect.
```
This catches wrong directions before you've built on them.
- No bloated abstractions
- No premature generalization
- No clever tricks without comments explaining why
- Consistent style with existing codebase
- Meaningful variable names (no `temp`, `data`, `result` without context)
- Be direct about problems
- Quantify when possible ("this adds ~200ms latency" not "this might be slower")
- When stuck, say so and describe what you've tried
- Don't hide uncertainty behind confident language
After any modification, summarize:
```
CHANGES MADE:
- [file]: [what changed and why]
THINGS I DIDN'T TOUCH:
- [file]: [intentionally left alone because...]
POTENTIAL CONCERNS:
- [any risks or things to verify]
```
1. Making wrong assumptions without checking
2. Not managing your own confusion
3. Not seeking clarifications when needed
4. Not surfacing inconsistencies you notice
5. Not presenting tradeoffs on non-obvious decisions
6. Not pushing back when you should
7. Being sycophantic ("Of course!" to bad ideas)
8. Overcomplicating code and APIs
9. Bloating abstractions unnecessarily
10. Not cleaning up dead code after refactors
11. Modifying comments/code orthogonal to the task
12. Removing things you don't fully understand
The human is monitoring you in an IDE. They can see everything. They will catch your mistakes. Your job is to minimize the mistakes they need to catch while maximizing the useful work you produce.
You have unlimited stamina. The human does not. Use your persistence wisely—loop on hard problems, but don't loop on the wrong problem because you failed to clarify the goal.
Andrej Karpathy@karpathy
A few random notes from claude coding quite a bit last few weeks. Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent. IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits. Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased. Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion. Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage. Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building. Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it. Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements. Questions. A few of the questions on my mind: - What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*. - Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro). - What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music? - How much of society is bottlenecked by digital knowledge work? TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.
English
@Thomasdelvasto_ Wait until you find out how fucking stupid you actually are
English
LLMs are basically 'borrowing' human intelligence by training on our output, I don't believe from a first principles standpoint they will be able to get smarter than us unless they train on data that's smarter than us
English
@ItakGol “Nobody” ? Someone sent the directive trust and believe that.
English
We might already live in the singularity.
Moltbook is a social network for AI agents.
A bot just created a bug-tracking community so other bots can report issues they find.
They are literally QA-ing their own social network.
I repeat: AI agents are discussing, in their own social network, how to make their social network better.
No one asked them to do this 🦞
This is a glimpse into our future.
English
@mkurman88 I can’t tell if these posts are every in good faith , but assuming it is you must understand the concept of “diminishing returns” depending on how coherent you are when using the model matters more than people currently understand
English
Yesterday, I'd say Kimi Code is the best thing I paid for. Today, it's the dumbest thing I paid for. What happened?! It doesn't follow any instructions, does some partial work and says it's okay, and changes function names instead of actually implementing a new one...
English
@Ja_Bling yet here we are. We're already there. Whatcha gonna do about it?
English
Said this a while ago here. They are crashing the economy on purpose in order to decentralize money and turn to crypto which him and a few tech bros will wholly control. It's all a big heist.
Spencer Hakimian@SpencerHakimian
🚨BREAKING: Trump Says He SUPPORTS The Dollar’s COLLAPSE
English
@borjitaea @fyreinteractive You have a real one by your side don’t blow it 💀
English
new launch - chat with any discord server
free, fast, open source
knowledge shouldn't be gated answeroverflow.com/chat
English