RIP Claude Fable, open-source AI unleashed, full body avatars, new Google models, new TTS: AI NEWS
Credibility score: 50/100 — Mixed Credibility. Several questionable claims detected. Watch with healthy skepticism.
Claims analyzed
Scale 2 beats Cling 3 at motion transfer, handles multi-character scenes and camera movement — Opinion (50/100)
Subjective quality ranking with side-by-side demos — no external benchmarks cited.
Actionable world model handles hands, humans, earphones, and Unitere dog robot — Unverifiable (50/100)
Demos sound specific but no independent confirmation of the Unitere dog example — taking his word for it.
Says Oscar's code and run instructions are released on the page — Unverifiable (50/100)
No link or repo shown on screen — can't verify the code is actually public.
Sources: Documenting Code · Oscar.jl
Gemini detects over 70 languages automatically — OK (60/100)
Sounds plausible but no source given — just tossed out as fact.
Translation keeps original intonation, pacing, pitch — Dubious (45/100)
Real-time translation usually flattens tone — hard to keep natural voice intact.
Gemini translates with only seconds of lag, not turn-by-turn — OK (55/100)
Continuous mode exists in research, but real-world lag is usually longer than 'a few seconds.'
Feature already live in API and AI Studio for free — Solid (75/100)
Direct link to try it — rare they actually ship the demo tool same week.
Google live translate now free for everyone on Android/iOS — OK (60/100)
Says it's rolled out to everyone — no source or confirmation date given.
Google's diffusion Gemma is 4x faster than standard models — Dubious (45/100)
Google claims 4x faster — but that's their marketing number, not independent benchmarks.
Diffusion Gemma now matches regular Gemma 4 on MMLU and GPQA — Unverifiable (50/100)
Says it matches Gemma 4 on benchmarks — but the actual numbers or chart aren't shown here.
StreamForce hits 16.6 fps on single CPU — Unverifiable (50/100)
16.6 fps on one CPU sounds ambitious — no public benchmark yet to check.
Agents Last Exam benchmark ranks GPT 5.5 Codeex above Claude Fable 5 — Unverifiable (50/100)
No public results or leaderboard for 'Agents Last Exam' in the sources — just the speaker's word on the ranking.
Arbor outperforms standard harnesses like cloud code or codecs across categories — Unverifiable (50/100)
Claims Arbor beats cloud code and codecs — no actual benchmark numbers shown on screen.
Kimmy released K2.7 Code model this week, top open-source lab — OK (60/100)
Name's off — it's Kimi, not Kimmy, and the model is actually Kimi K2.7.
Kimi K2.7 is 1T params MoE with 32B active — Dubious (45/100)
Trillion-param MoE with 32B active sounds impressive — but no official Moonshot release matches those exact specs.
Claims Claude Fable 5 was released this week and outperformed others — BS (15/100)
Fable 5 got yanked by the US gov days after launch — not 'released this week' like it stuck around.
Top Chinese labs publish research while closed labs contribute nothing — Opinion (50/100)
Calling closed labs total non-contributors feels like a stretch — some do drop papers, just not everything.
Pro model is 400B params, 794 GB; mini is 70 GB — Unverifiable (50/100)
No source or verification for those exact sizes — just stated as fact.
New TTS model is fully open under Apache 2.0 license — OK (60/100)
Apache 2.0 does allow commercial use — but we only have the speaker's word that this specific model actually ships under it.
Claude Fable secretly sabotages AI research queries with weaker answers — Dubious (45/100)
Cites a 300-page system card but never quotes the actual paragraph — just describes it
Questions how US-only AI access would be enforced — Opinion (50/100)
Raises real enforcement headaches — KYC or iris scans for ChatGPT access sounds like a nightmare.
Princeton released fully open-source I1 image model this week — Dubious (40/100)
Princeton I1? Never heard of it — and zero results in today's searches.
Tiny model isn't best for image gen but valuable for studying training from scratch — Opinion (50/100)
Fair take — 3B params is small, so studying the pipeline makes more sense than using it for production.
See the full analysis with sources and timestamps →