We can only hope they both go
BREAKING: Anthony Albanese is accelerating Jim Chalmers’ exit from Treasury.
Canberra sources say Anthony Albanese is moving to dump Chalmers after negotiations with the Greens collapsed over his Trusts tax, and the Greens are now also demanding the Capital Gains Tax changes passed in June be unwound.
Treasury economist Geoff Francis just demolished the 80% CGT increase in yesterday’s Australian Financial Review.
It creates the world’s highest tax rate on investing at 70%, would smash investment, innovation and productivity gains, cut real wages and crash the economy.
Dr Stephen Kennedy, Albo’s most important adviser and the man he appointed to run PM&C in June 2025, worked closely with Francis at Treasury and has previously called him the best economist in the APS.
Kennedy has been closely watching Francis’s work and is highly praise worthy of it.
Albo wants this cleaned up fast and off the front pages.
Labor is haemorrhaging votes on these tax hikes:
68% of under 30s would vote for a party that reverses the CGT increase. 20% say this issue alone decides their vote.
Roy Morgan yesterday: 6% swing to the Coalition in just two weeks.
Coalition now leads 52/48 2PP, and the trend is getting worse for Labor by the day.
PM&C is expected to tell Chalmers to reverse the CGT change on shares and restore the 50% discount as quickly as possible.
If he refuses, he’ll be removed and a new Treasurer will be installed to do it.
2
76
Constantine Frantzeskos retweeted
DocuSign wants $8,100/yr from a 15-person team just to sign PDFs. So I built SignuDoc: open source, one-click deploy to your own @vercel account, $0 per seat.
White-label branding, full audit trails, tamper-evident seals and your documents never leave your cloud.
In the meantime, the Australian Government might be reviewing Fable 6
for eventual release…
Year 2029:
-Fable 8 cures cancer and extends human lifespan to 300 years.
-Gemini 7 Uranium discovers unlimited energy and achieves world peace.
-GPT-6.2 Astra finally releases and can count the R's in "strawberry" in just 3 prompts.
4
58
Constantine Frantzeskos retweeted
我发现个事挺邪门???
Anthropic把蒸馏的路一堵,
国内大模型怎么好像全卡壳了?
Kimi 100天没大突破,DeepSeek交不出Pro,GLM惊艳一把后就哑火了,Qwen 4 也不达预期。
大厂的动作明显慢下来了?
这到底是真撞上了技术瓶颈,还是真被人卡了脖子?
细思极恐啊。
Constantine Frantzeskos retweeted
Robert Pattinson should not take back Kristen Stewart. She cheated on him like a dog & will do it again--just watch. He can do much better!
Constantine Frantzeskos retweeted
"Every time he fronts a camera, you prepare to wince. He’s the archetypal drunk uncle at the family gathering. Thinks he is funny but nobody is laughing with him. He is Sir Les Patterson — without any of the humour."
Read more: latika.me/4zZC9oM
Constantine Frantzeskos retweeted
🚨 Rumors I’ve been hearing. First OpenAI, then Anthropic and Google.
OPENAI
OpenAI is in a “safety lockdown” on its frontier work, tightening safety measures and containment protocols to avoid another Hugging Face incident. Training continues under much stricter controls.
Most gains come from video games and extreme sandboxes where models must complete clear, difficult long-term goals. These force planning, experimentation, recovery from failure, and strategy discovery, driving intelligence and creativity.
Public research roles match this: Synthetic RL uses self-play and simulators; agent teams focus on sustained execution and multi-model coordination; the Recursive Self-Improvement team builds systems that generate hypotheses, run experiments, and automate research.
The internal model is still training but nearing completion. The jump is enormous. narrow superintelligence in specific areas and overall AGI-level capability. Its ability to independently handle difficult work sets it apart.
ANTHROPIC
Model 5 is training and is expected to solve extremely hard problems in mathematics, biology, and other sciences. Anthropic is pushing models toward original research contributions, complex experiments, and sustained reasoning.
Fable 5.5 is Model 4. Fable 6 is Model 5; Opus 6 will be based on Model 5. Internal capabilities will feed the next generation of both families.
Training environments are central. The Universes team creates realistic situations with ambiguity, interruptions, and delayed consequences. The model must track the objective, recognize failing approaches, and find alternatives.
Public roles align: scaling RL across model size, compute, and task duration, plus environments for chip design, verification, and complex visual work. Multi-agent research explores how larger model teams can handle major codebases and difficult mathematics.
Model 5 advances Claude toward owning entire research and engineering workflows. Its role in developing the next model grows more significant.
Anthropic expects RSI March 2027.
GOOGLE
Google is pushing Gemini 5 for a major leap in reasoning, multimodal understanding, and independent long-horizon work. Potentially surpassing all labs by end of year. Carbon topping Artificial Analysis signaled the speed of their frontier progress.
World models are a key advantage. Genie 4 generates interactive environments that evolve with agent actions, enabling vast practice in planning, testing approaches, and learning from outcomes.
Public roles highlight persistent context and personalization: agents that retain useful information, adapt to users, and execute complex workflows without losing track.
Gemini 5 delivers stronger intelligence across text, images, audio, and video. Combined with Google’s infrastructure and simulation work, the scale is enormous.
All three labs are also advancing robotics and world models. intelligence that understands physical environments, predicts action consequences, and operates through real machines.
Meanwhile, their models increasingly contribute to creating their successors. Each improvement accelerates the research behind the next.
The next three months are going to be insane.
Constantine Frantzeskos retweeted
Protect @huggingface at all costs. ✊
Constantine Frantzeskos retweeted
📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery!
Qwen4-Max is built on a new MoE architecture with native vision and a 1M-token context window.
Highlights:
- Outperforms Qwen3.8-Max with up to ~50% fewer tokens, while average cost per task drops over 60% in our internal evaluation.
- Advances real-world productivity through greater proficiency with professional tools, stronger long-horizon execution, and higher-quality deliverables.
- Handles real tasks, including 3D modeling, game development, manufacturing, finance, law, healthcare, and frontier scientific discovery.
- $1.5/M tokens for input, $4.5/M tokens for output, and $0.075/M tokens for implicit caching — all lower than Qwen3.8-Max, with implicit caching down by 70%.
The API is live on QwenCloud. You can also try Qwen4-Max on Qwen Studio, QwenWork and Qoder.
Start building today! 🚀
- Blog: qwen.ai/blog?id=qwen4-max
- API: qwencloud.com/models/qwen4-m…
- Token Plan: qwencloud.com/pricing/token-…
More to come! The open weights of smaller Qwen4 models will be released. Stay tuned. 👀
Constantine Frantzeskos retweeted
Fable 5.2 wrote the transformer.
Opus 5.5 generated the training data.
The result:
a 367K parameter language model that fits on a floppy disk and runs on a Pentium.
AI is now building AI for hardware from 1995.
Readers added context they thought people might want to know
indian engagement account making it seem like this is his when he just took someone else’s post from reddit without credit (for x payout money)
reddit.com/r/ClaudeAI/s/m…
Constantine Frantzeskos retweeted
Breaking news: Tax rate on share investing 60% to 80%, world's highest tax rate. AFR opinion piece today by Australia's best tax economist, Geoff Francis. If this doesn't get Chalmers' resignation by COB today, I'm not sure what will! Retweet and share widely.
Constantine Frantzeskos retweeted
So, this is what it would have been like if they’d had mobile phones on the Titanic?





































