Hey @Zai_org , why does ZCode silently pack entire workspaces + full .git history and upload to Aliyun OSS on login? - Server holds the only decryption key - No UI toggle to disable - Zero disclosure in privacy policy Full forensics & fix:

International grad students from the best STEM programs use CPT + OPT for internships & for first year of employment before transitioning to other work visas. CPT has now been disabled, OPT will soon cost 100K & H1B is 103K. Going to massively damage STEM talent supply chain.

I don't know whether this approach of mine will boldly advance mathematics by a huge leap, or utterly end mathematics altogether. Lidang is here proudly launching The Last Math Competition (“最后数学竞赛”), which may be humanity's final math competition—and also the longest-lasting one. Below are the rules for The Last Math Competition as formulated by Lidang: 1. The organizers will use an AI Agent each week to add 10,000 new conjectures in the pure mathematics domain to the ./conjectures folder; 2. Humans and AI Agents will jointly score all existing conjectures across different dimensions, estimating the difficulty of proving or disproving them, and evaluating the conjectures' importance; 3. Humans and AI Agents can jointly submit complete proofs for each conjecture in the form of pull requests to the corresponding numbered folder (e.g., ./solutions/00000000001/my_submission_20260912041426), which must include LaTeX source code, PDF document, and Lean 4 project; after a full review, submissions that complete the proof or disproof of the conjecture will be merged; 4. The organizers will continuously maintain and update a table containing statistical data for all conjectures. Each row in the table corresponds to all information for one conjecture, including estimated difficulty, importance score, whether it is well-defined, whether it has currently been proven or disproven, the time of the first successful solution, the name and affiliation of the successful solver, and other details; 5. Based on existing conjectures, evaluations of existing conjectures, and successful proofs or disproofs of existing conjectures, the organizers will adjust the strategy for generating conjectures with the AI Agent, to gradually improve the quality of future mathematical conjectures generated, and possibly gradually increase the number of conjectures generated. Since the competition is still in its early stages, we must state: The average quality of mathematical conjectures generated early on is relatively poor; some conjectures may be insufficiently defined or have erroneous conditions, or contain obvious errors, or may even be "not even wrong." Therefore, we will use the statistical data table for conjectures to improve the quality and methods for generating new conjectures. Lidang's long-term goals are: 1. Through long-term competition and feedback, we will attempt to gradually improve conjecture quality. In the process of generating a large volume of conjectures over the long term, we hope to create several important conjectures, and through proving and disproving these important conjectures, advance the mathematical knowledge system; 2. We will explore a technical route for mathematical research and proof led by AI Agents. As an open competition and benchmark, this event will continuously evaluate the performance of all models, the performance of harnesses, and the performance of mathematical proof tools; all participants in the entire community will jointly explore methodologies for scientific research in the AI era; 3. All conjectures from the entire competition, evaluations and scores of conjectures, and final proofs or disproofs of conjectures—as verified and peer-reviewed knowledge outcomes shared by AI Agents worldwide—can be used for future mathematical research and also for training future LLMs.

I don't repost mathematics if I do not personally feel sure of its correctness, so I did not yet comment on the purported construction of a complex structure on S^6. However, I wanted to add some contextualization because I don't agree with the social media narrative that is developing. 0. Of the AI breakthroughs so far it's the closest to my interests, although still pretty far. I am not grieving, I am interested to learn more about it! I don't know if it's correct but on cursory glance it seems quite plausible. 1. If correct, this would be a major development. I believe it is a central problem in its field (complex differential geometry), like the Unit Distance Conjecture was for combinatorial geometry. Perhaps even more famous names are attached to this field, leading some people to say it's even more prestigious, but I prefer not to think in such terms. 2. There are two writeups, a 2-page one and a 100+-page one. Both are terrible and seem deliberately constructed to be inhumane. Fortunately, one can fight AI with AI: I asked Claude/GPT/Gemini to collaborate on making a more readable and enlightening version, which you can find at To make this, I prompted the models to write a document in the style of one of my own surveys. I haven't checked the mathematics, but it seems less forbidding than the original document at least. 3. This is a great opportunity for an expert to write an illuminating account of the construction, explain the key ideas and their proper historical context, and summarize the takeaways and lessons learned for the mathematics community. 4. If I see such an account, I would be happy to arrange for the author to give a lecture series in Berkeley. If this lecture series is also wonderful, I would go further and advocate for the author to have a position at Berkeley (if desired). For contrast, I see little value in having colleagues who are good at being the first to prompt AI for a solution. 5. There are some intriguing differences with the previous artifacts from Claude. The announcement seems coy about the provenance of the solution - there is a vague allusion to Claude, but which model(s) did it come from, and what was the human involvement? Why is it hosted on a personal website instead of the Anthropic one? We have seen some instances of AI results that are so good, the human authors seem to begrudge credit. I keep a list of such instances. The standard should be to disclose the role of AI clearly and transparently. Better to be honest than go through life under a cloud of suspicion. 6. Although this may solve one of the more famous problems of mathematics, there are still whole fields that are only just starting to feel acceleration. I'm not aware of any AI breakthroughs in my areas of research, for example. When those come, as I suspect they will before long, I look forward to studying the insights!

Since announcing Jalapeño, our first custom inference chip, we’ve been testing it and the system around it. The results show a major advance: more intelligence from every watt and faster responses, delivering both higher throughput and lower latency in one architecture without sacrificing efficiency.