AI Engineering Signal #37
Thinking Machines ships TML-Interaction-Small 276B-A12B, a purpose-built voice interaction model that eliminates the need for separate voice activity
Signals
Thinking Machines ships TML-Interaction-Small 276B-A12B, a purpose-built voice interaction model that eliminates the need for separate voice activity detection
realtime voice AI now operates end-to-end with lower latency and better turn-taking.
Latent Space
TabPFN-3 released
pretrained tabular transformer now handles 1M rows, directly competitive with gradient-boosted trees on medium-scale problems.
MagicQuant v2.0
hybrid GGUF quantization with Unsloth dynamic learned configs extends the Pareto frontier of size vs. quality for local models.
Google and SpaceX discuss orbital data centers
moving compute off-planet could bypass land, power, and cooling bottlenecks that already cap training clusters.
TechCrunch
Berkeley researchers find new pathway to energy-efficient chips
discovery could cut AI inference power requirements, a hard constraint on deployment scale.
Web
Anthropic-SpaceXai sign 300MW/$5B/yr Colossus I compute deal
training infrastructure spending continues to grow exponentially, even as others cut headcount.
Latent Space
The Take
Voice is leaving the research demo phase — dedicated models now do the whole stack — while compute infrastructure goes extraterrestrial and quantization tricks make small models viable. The split between "more compute" and "smarter tradeoffs" deepens.
Subscribe
Related Signals