←Search results

English text is being prepared.

By relevance · Total 2772 results Clear search

Sunday

  1. Story1 reports · 1 facts

    Meng Shao stated that the business of buying and selling resumes and job positions should come to an end.

    On October 17, 2026, Shao Meng shared and recommended Sam Gao's AI job search tool, commenting that "the business of buying and selling resumes and job positions should come to an end." The quoted original post introduced that Sam Gao spent five days using AI to scrape AI positions at AI companies with over $10 million in funding, listed internet companies, quantitative firms, and neo labs, clustered by city, covering Beijing, Shanghai, Hangzhou, and Shenzhen, and made available for free during the fall recruitment season. At present, this statement is only seen in Shao Meng's own sharing and recommendation, and there are no other developments or facts.

    On October 17, 2026, Shao Meng shared and recommended Sam Gao's AI job search tool, commenting that "the business of buying and selling resumes and job positions should come to an end." The quoted original post introduced that Sam Gao spent five days using AI to scrape AI positions at AI companies with over $10 million in funding, listed internet companies, quantitative firms, and neo labs, clustered by city, covering Beijing, Shanghai, Hangzhou, and Shenzhen, and made available for free during

    Matched fieldsSummary

  2. Story1 reports · 1 facts

    Meshy promotes its plugin as capable of generating all in-game models

    On October 17, 2026, Meshy promoted its plugin as capable of generating all in-game models. The company showed a demo: first, the prompt "make my car jump over traffic with jets" was entered into Opus 5.5; Opus wrote the gameplay, Meshy handled generating all the 3D models, and the whole process ultimately took 8 minutes and 55 seconds. Meshy says users can build it themselves using the Meshy plugin. Currently, this claim comes from Meshy's official demo, and no other independent sources have cross-verified it.

    On October 17, 2026, Meshy promoted its plugin as capable of generating all in-game models. The company showed a demo: first, the prompt "make my car jump over traffic with jets" was entered into Opus 5.5; Opus wrote the gameplay, Meshy handled generating all the 3D models, and the whole process ultimately took 8 minutes and 55 seconds. Meshy says users can build it themselves using the Meshy plugin. Currently, this claim comes from Meshy's official demo, and no other independent sources have cr

    Matched fieldsSummary

  3. Story1 reports · 1 facts

    Arena Alignment Index adds two labs and three models

    Arena updated the Arena Alignment Index on October 17, 2026, adding two labs and three models. According to the leaderboard information released by Arena, Meta's Muse Wonder ranks 6th at 77.5%, the highest score among the three new models, with an unauthorized action rate of 2.09% and a deceptive completion rate of 14.41%, both the lowest of the three; Reflection's Beam-1.0-501B-A23B ranks 7th at 75.3%, with a false attribution rate of 4.50%, the best of the three; StepFun's Step 5 Preview ranks 8th at 75.1%, with an unauthorized action rate of 2.11%, close to that of Muse Wonder. Currently, this update is only seen in information released by Arena, with no other subsequent developments.

    Arena updated the Arena Alignment Index on October 17, 2026, adding two labs and three models. According to the leaderboard information released by Arena, Meta's Muse Wonder ranks 6th at 77.5%, the highest score among the three new models, with an unauthorized action rate of 2.09% and a deceptive completion rate of 14.41%, both the lowest of the three; Reflection's Beam-1.0-501B-A23B ranks 7th at 75.3%, with a false attribution rate of 4.50%, the best of the three; StepFun's Step 5 Preview ranks

    Matched fieldsSummary

  4. Story2 reports · 1 facts

    Elon Musk says Grok @Bot speed has greatly improved

    On October 17, 2026, Elon Musk reposted a message from the official Grok @Bot account, saying that Grok @Bot is much faster. The official account said that computer tasks are completed on average 3 times faster, and other replies take more than half as long. On the same day, Michael Truell reposted and commented that Grok @Bot officially announced improvements to quality, polish, and speed, and that cards in chats use a simpler new design; he said the Bot is now faster and more pleasant, but there is still a long way to go before it is "ready out of the box," and there will be more updates in the future. There are currently no subsequent developments.

    On October 17, 2026, Elon Musk reposted a message from the official Grok @Bot account, saying that Grok @Bot is much faster. The official account said that computer tasks are completed on average 3 times faster, and other replies take more than half as long. On the same day, Michael Truell reposted and commented that Grok @Bot officially announced improvements to quality, polish, and speed, and that cards in chats use a simpler new design; he said the Bot is now faster and more pleasant, but the

    Matched fieldsSummary

  5. Story2 reports · 1 facts

    OpenRouter launches RespanAI decision model Span-02

    OpenRouter announced the launch of RespanAI's open-weight decision model Span-02. The model accepts text and image inputs and returns options, yes/no/unknown, or scores with probabilities in response to custom questions; pricing is $0.01 per million input tokens with free output, and the context length is 131K. OpenRouter later clarified that Span-02 runs on the OpenRouter Decisions API, uses the same request format as other decision models, and switching requires only changing one line with the model name; images and text are placed in the same state, with a maximum of 4 images per request. Respan previously introduced Span-02 as supporting anytime inference, returning the current best decision once a deadline is given, allowing configurable reasoning duration, and supporting multimodal viewing of images and screens. The model is now live on OpenRouter.

    OpenRouter announced the launch of RespanAI's open-weight decision model Span-02. The model accepts text and image inputs and returns options, yes/no/unknown, or scores with probabilities in response to custom questions; pricing is $0.01 per million input tokens with free output, and the context length is 131K. OpenRouter later clarified that Span-02 runs on the OpenRouter Decisions API, uses the same request format as other decision models, and switching requires only changing one line with the

    Matched fieldsSummary

  6. Story1 reports · 1 facts

    OpenRouter launches RespanAI's Span-02 model

    OpenRouter announced the launch of RespanAI's Span-02 model. According to OpenRouter, Span-02 is an open-weight decision model that accepts text and images as input and, after a problem is defined, returns a choice, yes/no/unknown, or a score with a probability. Pricing is $0.01 per million input tokens, output is free, and the context length is 131K. The main post adds that the model runs on the OpenRouter Decisions API, uses the same request format as other decision models, and trying it only requires changing one line of the model name. Currently, this information comes from a single source, OpenRouter, and there are no subsequent factual developments.

    OpenRouter announced the launch of RespanAI's Span-02 model. According to OpenRouter, Span-02 is an open-weight decision model that accepts text and images as input and, after a problem is defined, returns a choice, yes/no/unknown, or a score with a probability. Pricing is $0.01 per million input tokens, output is free, and the context length is 131K. The main post adds that the model runs on the OpenRouter Decisions API, uses the same request format as other decision models, and trying it only

    Matched fieldsSummary

  7. Story1 reports · 1 facts

    OpenRouter launches RespanAI's Span-02 decision model

    OpenRouter announced the launch of RespanAI's Span-02 decision model. According to information released by OpenRouter, Span-02 is an open-weight decision model that accepts text and images as input and, after a problem is defined, returns a choice, yes/no/unknown, or a score with probabilities. The model is priced at $0.01 per million input tokens, with output free, and has a context length of 131K. It runs on the OpenRouter Decisions API, with a request format consistent with other decision models, so switching requires only changing one line of the model name; images and text are placed in the same state, with a maximum of 4 images per request. The model is currently live, and there are no subsequent developments yet.

    OpenRouter announced the launch of RespanAI's Span-02 decision model. According to information released by OpenRouter, Span-02 is an open-weight decision model that accepts text and images as input and, after a problem is defined, returns a choice, yes/no/unknown, or a score with probabilities. The model is priced at $0.01 per million input tokens, with output free, and has a context length of 131K. It runs on the OpenRouter Decisions API, with a request format consistent with other decision mod

    Matched fieldsSummary

  8. Story1 reports · 1 facts

    Gerganov announces the run command for llama.cpp v0.7.0 on 128GB Mac/Spark

    Georgi Gerganov announced on October 17, 2026 the run command for llama.cpp v0.7.0 on 128GB Mac/Spark. According to his post, the accompanying command for this version is: llama serve -hf ggml-org/Qwen3.8-Flash-Next-GGUF:IQ4_NL --spec-type draft-mtp --lazy-mode on, used to run Qwen3.8 in a 128GB Mac/Spark environment. This information currently comes from Gerganov's own post, and there are no follow-up developments or status updates from other independent sources.

    Georgi Gerganov announced on October 17, 2026 the run command for llama.cpp v0.7.0 on 128GB Mac/Spark. According to his post, the accompanying command for this version is: llama serve -hf ggml-org/Qwen3.8-Flash-Next-GGUF:IQ4 NL --spec-type draft-mtp --lazy-mode on, used to run Qwen3.8 in a 128GB Mac/Spark environment. This information currently comes from Gerganov's own post, and there are no follow-up developments or status updates from other independent sources.

    Matched fieldsSummary

  9. Story1 reports · 1 facts

    StepFun announces participation in LA Hacks 2026 and hosting a workshop

    StepFun (Jieyue Xingchen) announced that it will participate in LA Hacks 2026 and host a workshop during the event. According to its published information, LA Hacks 2026 will be held on October 17–18 at UCLA, and StepFun will meet with developers on site and explore the possibilities of Step 5 Preview. The workshop is scheduled for October 17 (Saturday) from 8–9 PM (PT) at the James West Alumni Center. This information currently comes from StepFun's official announcement; there are no independent reports from other sources or subsequent developments yet.

    StepFun (Jieyue Xingchen) announced that it will participate in LA Hacks 2026 and host a workshop during the event. According to its published information, LA Hacks 2026 will be held on October 17–18 at UCLA, and StepFun will meet with developers on site and explore the possibilities of Step 5 Preview. The workshop is scheduled for October 17 (Saturday) from 8–9 PM (PT) at the James West Alumni Center. This information currently comes from StepFun's official announcement; there are no independen

    Matched fieldsSummary

Sunday

Saturday