#6217: Fine-Tuning a Trillion-Parameter Model at Home

What does it actually take to full fine-tune a trillion-parameter model? Sixteen terabytes of optimizer state, hundreds of GPUs, and a very big bill.

local-aiai-inferencefine-tuning

#6216: What Do You Call the Middle Tier Model?

Sonnet, Sol, Flash, Medium — the daily-driver models nobody can name, and why they're squeezed from above and below.

ai-inferencelocal-aiai-evaluation

#6215: Can Anyone Still Spot an AI Voice?

Clones now pass as human. Generic AI voices still don't. And the detectors? They fall apart in the wild.

speech-recognitiondeepfakesai-evaluation

#6214: Nobody Trains an Uncensored Model From Scratch

Every "uncensored" model is a renovation, not a new build — and the political lean lives in the agreement layer, not the refusals.

open-sourceai-alignmenttraining-data

#6213: Ten Weirdest AI Evals Ever Published

A pelican on a bicycle, a badger for Groundhog Day, and a benchmark where passing is the bad outcome — the strangest AI evals ever run.

ai-evaluationhallucinationsai-alignment

#6212: Woolly Sedate Snail and Other Model Failures

Most of Hugging Face's 2.96M models sit unread. But the dust has a personality — snails, pirates, and a matchbox that resigns.

open-sourceai-evaluationhallucinations

#6211: Attention Is Still Winning: The Hybrid Model Census

Mamba layers shipped — inside hybrids. A census of Qwen, Nemotron, Kimi, Granite, and why MiniMax went back to full attention.

ai-inferencelocal-aitransformer-architecture

#6210: Who Decides If an Election Is Real?

57 countries hold multiparty elections that aren't real contests. So who certifies a vote — and why do clean democracies skip full scrutiny?

electionshuman-rightsinternational-law

#6209: Watching Elections: Who Collects the Numbers, and Why

From CIA tasking orders on France's 2012 race to hedge funds buying private Brexit polls — who watches elections, and what they do with the numbers.

intellectual-propertyelectionssignals-intelligence

#6208: Israel's 120-Seat Forecasting Ministry

A windowless office, a serious budget, and one job: forecast Israel's 120-seat Knesset as well as it can be done.

israelelectionsforecasting-and-wargaming

#6207: Two WERs, One Model: Inside ASR Eval Harnesses

Same model, same audio, two different word error rates. The gap lives in the normaliser — and that's just layer two of three.

speech-recognitionai-evaluationlocal-ai

#6206: Monte Carlo: How Random Guesses Solve Impossible Problems

From solitaire at Los Alamos to neutron diffusion, why rolling dice beats careful grids — and how many runs is enough.

statisticstaxonomy-and-classificationhistory-of-technology

#6205: Where Should Your Notebook Actually Run?

Colab's free GPU comes with a 12-hour wall and hidden secrets in your saved outputs. Here's where notebooks should actually live.

open-sourcecloud-computinglocal-ai

#6204: Two Humans, One Table Each: Calling Israel's Election

No models this time — just two seat-by-seat tables, fifty vs. fifty-nine, and a defector hunt two seats short of a majority.

israelelectionsforecasting-and-wargaming

#6203: How Claude Code Sessions Message Each Other

Two terminals, one machine: how Claude Code sessions find each other by name, pass messages, and subscribe to an idle ping.

ai-coding-assistantsmodel-context-protocolmulti-agent-systems

#6202: Israel Votes, Part 6: Asking the Decision Models

Two decision models, identical evidence, one table of betting prices — and both flip their pick for Israel's next leader.

israelai-evaluationforecasting-and-wargaming

#6201: Inside Cloudflare's Clef: A Language Model That Can't Speak

Cloudflare's Clef reads text, images, and video but outputs only probabilities — no words. Here's what's actually under the hood.

open-sourceai-inferencelocal-ai

#6200: SekerKolshehu: seats behind the seat count

A new open source tool simulates 100,000 election allocations from a single poll — and shows why "4 seats" is less certain than it sounds.

statisticselectionsisrael

#6199: Running AI Experiments Before an Election

What MLflow, Braintrust and notebooks actually do when you have five days and one forecast to get right.

open-sourceai-evaluationworkflow-automation

#6198: Benchmarking a TTS Model on Your Own GPU

Real-time factor, warm-up artifacts, and why an 8-bit quantized speech model might not run any faster.

speech-recognitionlocal-aiai-inference

#6197: Talking to Your Own Voice Clone

What actually happens when you sit down and talk with an AI clone of your own voice — and what Evan Ratliff's Shell Game revealed.

speech-recognitiondeepfakestext-to-speech

#6196: Fine-Tuning ASR: When Your Model Memorizes You

Your fine-tuned speech model aces your training recordings — but does it actually work on fresh audio?

speech-recognitionai-evaluationlocal-ai

#6195: Me and My Voice Clone

Daniel talks to his own voice clone. He recorded his lines on a phone; the clone's lines are Meorav v2, a model trained on 88 minutes of his voice.

speech-recognitiondeepfakestext-to-speech

#6194: Teaching a TTS Model His Own Accent

One man fine-tuned a text-to-speech model on his own voice to get Irish-accented English that can say Hebrew words mid-sentence.

training-datatext-to-speechspeech-recognition

#6193: Why Email Images Still Phone Home

That "load remote images?" prompt is older than your phone. Here's what fetching a picture actually tells the sender.

digital-privacycyberattackssurveillance