china's new 5 year plan is expected to create 500,000 new jobs in AI while expanding vocational retraining programs and requiring data centers to hit carbon reduction targets using renewable energy but in the USA it's literally just um idk make better choices do u love israel

For legacy banks and orgs, it means rebuilding legacy infrastructure, changing processes, retraining people, and spending billions. blocking @AnthropicAI is much easier than rebuilding. Anthropic’s speed can’t be matched anyway. ⚡️

China just used the degree catalog as a continuous eval layer China's Ministry of Education revoked or suspended 12,200 undergraduate degree programs between 2021 and 2025 and added 10,200 new ones, touching more than 30% of the country's university programs. The 2026 catalog of undergraduate majors is the trigger for the next wave. Photography and traditional arts get culled. Embodied intelligence, intelligent imaging, and human machine collaboration get added. The whole catalog is being re-platformed on a 5 year cycle. The architecture is identical to what Adaline just shipped for AI agents. Production traffic is the eval signal. The catalog is the eval suite. New programs are the verified candidates that get human approval (in this case, the MoE) and ship back into the running system. The build is solved. The run is the problem. That applies to people now, not just agents. The pressure is the graduate jobs crisis. China is on track for 12.7 million graduates entering the 2026 market, many with degrees that the labor market is no longer paying for. The MoE is doing what any continuous eval layer has to do.. replacing the metric that is no longer load-bearing. The 4 year degree is now a worse predictor of the next 4 years than the catalog itself. The US, the UK, and Australia cannot move this fast. 4,000 institutions each setting their own catalog means the same speed of retraining is structurally impossible. China rebuilt the degree catalog. The West still has the same one. Embodied intelligence is the bellwether. Nine universities added new embodied AI majors this cycle. That is the same national drive as the May humanoid robots policy and the June OpenRouter Fusion panel (where two of the four budget models are Chinese). The degree pipeline, the industrial policy, and the model layer are all pointing at the same problem from three directions. The eval suite is the product. So is the degree catalog.

A. Santini et al. Rouen 🇫🇷 Epi2Diag building evidence for episignatures in NDD diagnosis. Episignature performance not uniform. e.g. 👍 Sotos, NIPBL (not other BafOpaththies) but some others less Retraining classifiers improved several weaker signatures. #ESHG2026

Anyway, one big takeaway is that predicting the jobs of the future is hard. Maybe the guys working in cyber will end up retraining as ballerinas. Or maybe we'll all end up in jobs which don't even have names yet. Read more here: observer.co.uk/news/business…

Maybe those hamsters should try brain retraining or being more positive and not fixate on the disease narrative 🤡

I see - so not using pre-trained weights, but rather fully retraining from scratch? If so, what data?

Wasn’t she a whore before she got married? Her husband wasn’t the problem. Her parents were. Until parents go back to retraining their daughter’s freedoms, nothing will change. This girl was very free from a very young age and it shows. Ofc she was sexually immoral.

The chart shows short-term pain from broader SaaS rotation AI fears, but the thesis is backwards. Stickiness is the moat in enterprise software. $NOW is deeply embedded in core workflows, data models, integrations, compliance and processes across IT/HR/CS/etc. Switching costs are massive (retraining thousands, data migration risk, downtime), that’s why renewal rates stay at 97-98%. AI agents with structured MCPs won’t replace it, it is embracing and enabling them. They have native MCP client server support so external agents (e.g. Claude) can securely interact with the platform. Their own Now Assist AI Agents and Otto are already agentic and act autonomously on real workflows. They just raised the 2026 AI revenue target to $1.5B. The moat isn’t a “floating dream” It’s widening as ServiceNow becomes the orchestration layer for AI agents, not the thing getting replaced. Fundamentals and adoption tell a different story than the chart and all the people selling the fear are the same will buy at double the price ‼️

Everyone talks about AI. Nobody talks about staring at training progress bars all weekend because you don't have decent GPUs. Got new custom data from one interested client and now retraining the model.

Upgraded IL-38, TU-142 and Midas. All of which which they operated earlier. Aren't they good enough? If not, why don't we hear anything in open against them like we do about LCA, LCH, Arjun, INSAS and all other local weapon systems? Is retraining, new logistics and doctrine not as big issues from shifting from Foreign1-to-Foreign2 as it from ForeignX-to-Indian?

What if AI retraining is just a comforting lie? japantimes.co.jp/commentary/…

EverOS 1.0.0 just dropped — and it's the most practical approach to agent memory I've seen this year. It's an open-source Python framework for self-evolving long-term memory that works across Claude Code, Codex, Hermes, and any other agent. One portable memory layer so context follows the work instead of staying trapped in one tool. The architecture is refreshingly simple: - Markdown as the source of truth — every memory is a .md file. Readable, grep-able, Git-versioned, opens in Obsidian. - Local stack: Markdown SQLite LanceDB. No MongoDB, no Elasticsearch, no Redis. - Dual-track memory: agent memory (cases/skills) and user memory (episodes/profile) extracted independently. - Multimodal ingestion: text, images, audio, PDFs, HTML, email — all unified into searchable memory. - Self-evolution: common skills extracted from usage patterns. Repeated workflows become reusable without retraining. The cleverest part: orthogonal retrieval. You can search independently by user_id, agent_id, app_id, project_id, and session_id. That means an agent working on Project A doesn't get confused by memories from Project B — a problem most memory systems don't solve well. Install is one line: `uv pip install everos` or `pip install everos`, then `everos init` and `everos server start`. OpenAI-protocol compatible — works with OpenRouter, vLLM, Ollama out of the box. This is the memory layer I'd build on if I were putting agents into production today. No cloud dependency, no vendor lock-in, and your memory is plain Markdown files you own. github.com/EverMind-AI/EverO…

subsequently, all young men were mobilized and deportation, about 30k to Siberia! it was called military retraining! All police, border guard, homeland defense, military groups etc.! It was actually a prison camp and about one third, of them died of hunger or disease in Siberia

Maybe retraining and up skilling is required

When things aren’t working or it doesn’t reach the 100% of intended behavior. Majority of people don’t think of running evals, which will burn tokens as well, and most aren’t even prepared to build one, and they don’t know how to verify evals. Then typically when the model reached 100% of its target on a particular task. Most people will push for more, 110% 120% and the incremental improvements between them can’t be explained if you don’t understand how transformers work or whatever architecture behind it (which is closed weights anyway) If you open the weights you can demonstrate at a lower level. But what most people see is magic tokens being chained around tool calls Fixing it externally requiring you to run evals which bring us back to the first problem. Retraining the weights, fine tuning require evals again

Another of my Favorites Sold on IWC! Retraining Your Brain iwe.one/VJlaV

TurboQuant (Google, ICLR 2026) compresses that cache to 3 bits per value. At least 6x less memory. Up to 8x faster attention on H100. No retraining, no accuracy drop. It rotates the data, then quantizes the hard part.