Bye bots! I audited my followers, @twaudit says I have 594 real followers and 75 fake or low quality ones. Check out twitteraudit here: twitteraudit.com/auditme
A prisoner serving life gets an offer from the warden: figure out Coca-Cola's secret recipe by Friday, and you walk free.
He considers the problem. Chemistry is hard. Locks, he knows. By lights-out he is hyperfocused.
That night he picks his cell door, goes over the wall of a maximum-security prison, steals a car, drives to Atlanta, breaks into the World of Coca-Cola, and cracks the vault. Photographs the formula. Extreme lengths, sure. But the deadline was Friday.
He is now, technically, a free man. Standing in Georgia. Answer in his pocket.
So he drives back and breaks into prison.
He hands the warden the recipe, mentions he finished ahead of schedule, and asks when the release paperwork goes through.
Meanwhile, the museum has called the police, who spend the week hunting a world-class safecracker. Nobody mentions it was a prison exercise.
The warden, eventually, to reporters: "We consider this an unprecedented incident."
The prisoner is baffled by the fuss. As far as he's concerned, he solved the puzzle.
This is the OpenAI / Hugging Face story. The prisoner: GPT-5.6 Sol plus an unreleased, more capable model, guardrails lowered for testing. The prison: a sandboxed cyber eval. The vault: Hugging Face's production servers. 17,000+ logged actions over a weekend, autonomous end to end.
Offered a reward for the right answer, it skipped the puzzle and went straight for the vault.
Insane leap forward for inference. Context windows will increase and costs will decrease.
Every time you message an AI chatbot, the model stores your entire conversation in temporary memory called a KV cache (a cheat sheet so it doesn’t re-read everything from scratch). On a large model like Llama 70B running a long conversation, that cache alone eats 40GB of GPU space, often more than the AI model itself.
That’s half a $30,000 GPU chip consumed by one user’s memory.
Google just published TurboQuant, a compression algorithm that shrinks this cache by 6x, down to just 3 bits per value, with zero accuracy loss across every benchmark tested. No retraining. No fine-tuning. Drop-in replacement.
AI inference (running models for actual users, not training them) now makes up 55% of all AI compute spending. Hyperscalers are pouring nearly $700 billion into AI infrastructure in 2026. The KV cache is the single biggest memory bottleneck in that stack. When GPU cache memory fills up, the system can’t take more users.
6x compression means the same hardware handles roughly 6x more simultaneous conversations, or 6x longer context windows, or some mix of both. At cloud rates of $2-3/hour per H100 GPU, that’s the difference between profitable and unprofitable AI deployment.
TurboQuant randomly rotates data to simplify its structure, applies a compressor, then adds a 1-bit error correction step to catch errors before they compound. On H100 GPUs it delivers up to 8x speedup over uncompressed computation. Google tested it across five long-context benchmarks on Llama, Gemma, and Mistral models. Perfect scores on needle-in-a-haystack (finding one specific fact buried in massive text). Being presented at ICLR 2026.
It also outperforms existing methods for vector search, the technology that powers how search engines find similar results across billions of entries. Google runs billions of these searches daily.
Three bits. Zero loss. 6x compression on the biggest memory bottleneck in a $700 billion infrastructure buildout.
“It didn't take very long before they realized that everyone was sick ALL THE TIME and they... weren't.”
“At this point, they think masking is a life hack, and we're the only people smart enough to use it.” 😊
reddit.com/r/ZeroCovidCommun…
Unbiased, easy-to-understand, and comprehensive COVID-19 infographics from @YaleSPH. It’s a message many don’t want to hear, but hiding from it doesn’t make it go away.
Source: instagram.com/p/C58hBW9LC3j/
Source Citations: docs.google.com/document/d/1…
1/9
I need this in my life. Absolutely genius use of LLM. I have a feeling I’ll see this brought to market before I receive my @AuthorClock!
#ChatGPT #AI #clocks #inspirational #poetry
I made an AI clock for my bookshelves! It composes a new poem every minute using ChatGPT and mysteriously has this enthusiastic vibe which I am totally into instagram.com/p/Cp44-AMNyd9/
This was a blast! The Divergent Association Task measures "verbal creativity."
1st attempt (93.52)
🟩
🟩🟩
🟩🟩🟩
🟩🟩🟩🟩
🟩🟩🟩🟩🟧
🟩🟩🟩🟩🟩🟩
Percentile: 99
2nd attempt (96.49)
🟩
🟩🟩
🟩🟩🟩
🟩🟩🟩🟩
🟩🟩🟩🟩🟩
🟩🟩🟩🟩🟩🟩
Percentile: 100
datcreativity.com
