极客邦科技创始人兼 CEO 霍太稳，为 1024 模力工场·AI 社区日的主论坛开场致辞，并以《AI 破局的生态法则：从零和博弈到共生智能》为题作主题分享。
霍太稳在开场致辞中表示，举办此次社区日、推动总部落地杭州，是希望立足本地，通过持续的社区建设与交流合作，为 AI 应用落地培育生态。
在主题分享中，他提出，AI 行业的关注点，正从大模型的参数竞争转向智能体及具体业务应用，于此同时，企业的需求正在变化：相比几天的培训，它们更希望服务方带着团队做出能用的智能体与 Skill，让学习直接连接实践。所以极客邦科技，也将“助力企业 AI 加速落地”作为业务升级的重要方向。
霍太稳表示，会用 AI 工具，不等于能用它解决业务问题。企业有需求、有预算，却不容易找到靠谱的团队；技术团队能做项目，也担心做完后收不到钱。模力工场想解决的，就是双方找人难、建立信任难的问题。
他介绍称，模力工场计划用 AI 帮助企业判断一个项目大概需要多少时间、多少预算，再根据团队过去的交付表现推荐合作方；技术团队也能据此了解企业需求和预算，选择合适的项目。双方建立合作后，平台还希望帮助跟进交付、评估结果。
他认为，这件事亟需更多生态合作伙伴一起来做，而极客邦科技过去近 20 年积累的企业与技术人才资源，正是推进这项工作的基础。
以下为霍太稳讲演的完整速记，模力工场在不改变原意的前提下进行了订正和编辑。
模力工场：连接 AI 供需，补上信任缺口
前两天说今天好像还有台风，我们还担心这个活动能不能举办。但好像杭州确实是一个福地，我一直认为它是一个福地，包括过去的几次台风，好像都完美地错过了。在我开场之前，简单说一下为什么要有今天这样的一个活动。
如果从大往小讲，第一个因素，肯定是从大的角度来看。前段时间，国务院还发了一篇关于促进新质生产力发展的文章。浙江省现在也在全力打造人工智能创新高地，杭州市也要争创人工智能创新发展第一城。我觉得，做事情还是要顺势而为。这是第一个因素。
第二个因素，我们都知道，要想促进行业或者产业的发展，生态是非常关键和重要的。所以，我们希望能够在杭州这个地方，把整个生态做得更好一点。
第三个，对于极客邦来讲，我们一直是以做社区、做学习平台为生的。要想把整个生态做好，这是一件长期的事情，需要脚踏实地、扎扎实实地做下来。
所以，这也是等一下我们要举行一个小仪式，极客邦科技南方总部、TGO 鲲鹏会全球总部落到咱们这个地方的一个非常重要的原因。我是去年 9 月份到杭州、到云谷中心来的，今年 6 月份正式启用办公室，但一直没有举行仪式，因为一直在忙。
前段时间，可能很多人关注过 GOAI 世界人工智能开源大赛，几十个人一直在忙这件事情。到了 10 月份，才算有那么一点时间。正好今天是 1024，也是咱们程序员非常喜欢的一个日子，就来做这样一件事情。
这是一个前情摘要。下面我再花十来分钟，给大家汇报一下目前整个极客邦，包括模力工场在做的一些事情。不讲太多的大道理，只谈我自己的一些小小看法。
现在行业公认的 AI 发展是分三个阶段的。如果大家关注一些 IT 报道或者 AI 新闻，大概率能够感受到，过去三年，尤其是去年之前，整个是大模型的天下，大家都非常努力地在拼各种各样的参数。但是，你们有没有感觉到，今年好像有一点降温？整个热度跑到哪里去了？有两个小例子，大家感受一下。
一个是原来 Teambition 的创始人齐俊元，他又去创业，做 Personal Agent，也就是个人智能体，应该叫 Today AI，现在整个估值是 10 亿美金。原来阿里出去的一位同学叫林俊旸，现在去做企业级智能体，整个估值是 50 亿美金。
所以，大家大概能够感受到，AI 正在进入千行百业，进入智能体这个阶段。为什么很多人称今年为“智能体元年”？我觉得大概和这个有一定关系。
当然，对于未来，也不敢做太多臆测，AI 发展得确实太快了。但在我看来，随着越来越多的智能体、Agent 出现，大概率会出现一个原生的 AI 操作系统，或者说，一个专门为智能体提供服务的操作系统。
当下的热点是，AI 怎么在千行百业里面用起来。所以，基于这样的判断，极客邦科技也把自己的愿景使命做了一个升级，里面加了一句非常关键的话，叫“助力企业 AI 加速落地”。
也就是说，在未来几年，我们主要的精力是帮助企业在内部，不论是工作流还是其他方面，能够用上 AI，看看能不能通过 AI 找到第二增长曲线，实现降本增效和提质。
当然，这并不是一个新鲜概念，很多企业也都在讲。包括我们公司，过去一两年也在这上面做了蛮多事情。比如极客时间在做 AI 落地案例集，把全行业里面比较有代表性的案例整理成册。如果大家关注极客时间，近期就能看到这个小册子。
我们也做了一个 AI 工程师知识库，里面包括大模型、智能体的应用。希望通过这样的方式，让大家了解整个行业在发生什么，有什么真正可以参考的案例。
如果你是一名 AI 工程师，怎么用大模型，怎么用这些工具，也能有一个参考。因为整个极客邦服务的，基本上还是这些 AI 工程师。大家如果感兴趣，可以直接扫描二维码。
对于企业服务，极客时间企业版是一个主力军。像现在大家比较关注的中国移动，几十万人的 AI 培训，未来四年都是极客时间在帮助他们做。包括华润集团 38 万人，从董事长到基层员工的 AI 培训，也是我们在帮他们做。
但是，我们从今年年初就发现，行业发生了一个比较大的变化。大家可能并不只是说，你过来给我做几天培训，而是希望你能不能带着我们的工程师、我们的团队，直接做出一些智能体、一些 Skill。
所以，基于这些判断，我们也把给企业提供的服务做了智能体化、Agent 化的升级，分成 AI 人才管理与发展、AI 培训与学习、AI 企业解决方案三大部分，用了差不多 8 个智能体，做了整体的改变。
这样，大家在企业内部进行技能提升、团队组织效率提升时，就更加偏向 AI 化和智能体化。
刚才说到的，主要是我们过去所做的一些业务，包括 InfoQ、极客时间的业务。但在我看来，这肯定远远不够。
InfoQ 提供的是信息服务，极客时间提供的是知识服务。我经常举一个例子：所谓的信息服务，就像你在一个大广场上，用大喇叭告诉别人，AI 很重要，智能体很重要，大家来听一听。
到了极客时间，更多的是，我知道这个技术很重要，我来教你学会它、使用它。但你学会了、会使用了，总还要有一个落地的场景。这就是模力工场希望达到的目的：真正成为一个企业 AI 落地服务平台。
如果用一句话来介绍，就是希望把企业的一些 AI 想法，变成真正的业务结果。
当然，你要改公司的愿景，或者把一个业务进行升级，这句话好讲，但真正把它落地、变成现实，挑战还是挺大的。
所以往回倒一倒，我们到底想解决什么问题？还是分两头来讲。
今年正好是极客时间 9 周年，过去 9 年里，服务了差不多 400 万个人用户、4000 多家企业。最近一两年，我也在走访很多企业，和他们打交道、交流。在这个过程中，发现他们有一个非常明显的痛点：我有预算，也有时间、有场景，但不知道到哪里找到一个靠谱的 AI 交付团队，帮我把问题解决掉。
他们最担心的是，我给了你时间，给了你钱，却没有拿到一个可落地的结果。
我自己也是创业者。对于企业经营来讲，最大的成本其实是时间。你做了一个错误的决策，但你的友商可能已经跑在前面，再去追赶，难度就比较大了。
对于很多 AI 交付团队，包括很多 FDE，其实也面临一个痛点：我懂 AI、懂技术，也有时间，愿意为你提供服务。但我担心的是，给你做了这些事情，能不能拿到你许诺的预算？
所以，即使互联网发展了这么多年，我们花了很多心思解决信任问题，但这个问题现在依然存在。我就在想，有没有可能通过模力工场，以 AI 的方式把这个问题解决掉？
大家可以想一想，一个企业有 AI 场景，有团队想为它提供服务，但双方可能都不太清楚，大概需要多少工期、多少预算才能完成。
如果有一天，企业在模力工场发布需求时，AI 供需大脑能够比较容易地告诉它，这个需求大概需要多少时间、多少预算才能完成；还能推荐平台上相对靠谱的团队，告诉它，这几个团队或者这几个人，过去交付的口碑不错，团队实力也不错，那效率是不是就能有比较大的提升？
对于入驻的 AI 团队或者超级个体，他们能够在平台上看到，哪些企业愿意花多少预算，解决什么问题，也可以选择信誉相对比较好的企业。
不仅是对需求进行智能匹配，在交付时，也能够对结果进行衡量。我觉得，这是模力工场接下来要花蛮多时间去做的事情，希望通过 AI，把双方连接在一起。
当然，很多人也会问，这样的业务形态并不是稀缺品，在互联网时代已经存在很久了，现在也有很多人在做。为什么极客邦还要参与进来，而且认为自己能够把这件事情做好？
我也列了几个原因。
首先，明年是 InfoQ 进入中国 20 周年，我们认为自己对技术、对 AI 还是相对比较了解的。
刚才也提到，如果加上 TGO 鲲鹏会、整个极客时间，我们过往服务了 400 多万个人用户、4000 多家企业。所以，我们认为自己对人才、对企业也是相对比较了解的。
你对 AI 了解，对人才了解，对企业了解，又积累了这些比较好的资源。所以，我们认为，对于极客邦、模力工场来讲，有责任站出来，扮演好这样的角色，完成这样的使命，真正让 AI 进入千行百业。
但这件事情，是不是只靠极客邦一家就能做好？我也不这么认为，尤其是在现在这样一个特别讲究生态的时代。所以，我们也开始严选一些生态合作伙伴，一起为企业服务。
这里简单列几个。
第一个是深空之序，它的创始人是前钉钉总裁叶军博士，花名不穷【“总裁”职务表述待核实】。今天，它的联合创始人，也是我们 TGO 鲲鹏会杭州分会的负责人芦宇峰，肉山，也来到了现场，等一下会做分享。
第二个是数势科技的李克峰博士，他是原来京东零售的 CTO。还有徐亚波，应该是教授，他背后的数说故事；包括桑文锋的神策数据，以及张新波的科瑞英特【人名及公司名待核实】。
列出的这几个生态伙伴，背后都有一个共同特点：它们都是我们 TGO 鲲鹏会的会员企业。
点到这里，大家大概就知道是怎么回事了。我们在利用极客邦科技过去 20 年积累的生态资源，把企业 AI 落地这件事情做好。
当然，也有很多小伙伴说，我们也想参与，应该怎么参与？非常简单，扫码参与。
开个玩笑。这里有两种情况：如果你有 AI 落地能力，希望在这里找到一些场景，可以像刚才提到的深空之序、数势科技、数说故事一样，成为我们的 AI 生态伙伴。
如果你是一家企业，有非常好的 AI 场景，希望找到比较靠谱的解决方案或者 AI 交付团队，也可以扫码参与。这是一个典型的双边平台。
我们希望和更多生态伙伴一起，把企业 AI 落地服务这件事情做得足够好。
最后，因为今天是 1024，再次祝所有程序员节日快乐！当然，现在叫 AI Builder，节日快乐！也希望大家今天学习快乐。
再次感谢大家的光临，谢谢！
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
The founder of Teletech and CEO Ho was able to open the main forum for the 1024 Modified Fields-AI Community Day and to share it under the theme " The Code of Ecology of the AI Buster: From zero to symbiotic intelligence ".
In his opening remarks, Houtha said that the organization of the Community Day and the promotion of the Landscape of Hangzhou at its headquarters were meant to foster ecology for AI applications through local and sustained community building and communication cooperation.
In the thematic sharing, he suggested that the focus of AI was shifting from LLM ' s parameter competition to AI agent ' s and specific business applications, and that, at the same time, the needs of enterprises were changing: compared with the training of a few days, they would have preferred service providers to bring a team to work with Skill, so that they could learn to connect directly to practice. As a result, Ultra-Hot State technology has also made “Assisting Enterprise AI Accelerating Landfall” an important direction for business upgrading.
Haugh said it would use an AI tool, not the same way it would solve operational problems. Businesses have needs and budgets, but they do not easily find reliable teams; technical teams are able to do projects and fear that they will not receive money after completion. The problem of finding and building trust between the two sides is the problem that the model workshops are trying to solve.
He described the model workshop plan as using AI to help enterprises determine how much time and budget a project might take and to recommend partners based on the team ' s past delivery performance; the technical team would also be able to learn about the enterprise ' s needs and budget and select the appropriate project. With the establishment of cooperation between the two parties, the Platform also hopes to help follow up on the delivery and evaluation results.
In his view, the matter urgently needed to be done by more ecological partners, and the business and technological resources that had accumulated over the past 20 years in Polar Science and Technology were the basis for taking the work forward.
The following is a complete schematic of the smooth presentation by Horten, which was revised and edited without changing the original intent.
Modular workshops: connect AI supply and demand and fill the trust gap
Two days ago we said there was a typhoon today, and we were worried about the event. But as if Hangzhou was a blessing, I always thought it was a blessing, including several past typhoons, and it seemed to have been missed perfectly. Before I begin, let me briefly state why there is an event like this.
The first factor, if from a broad perspective, is certainly from a broad perspective. During the previous period, the State Department also published an article on promoting the development of new quality productivity. The province of Zhejiang is also making efforts to build artificially intelligent innovation highlands, and Hangzhou City is also struggling to create the first city for artificially intelligent innovation development. I think we have to do things the way we do. This is the first factor.
The second factor, as we all know, is that ecology is critical and important if we are to promote industry or industry. So we want to be able to do the whole ecology better in this part of Hangzhou.
Thirdly, we have always lived as communities and as learning platforms for the extremist states. It is a long-term undertaking that needs to be done on the ground and solidly.
So that's one very important reason why we're going to have a little ceremony later on, where the Southern Teletech, the TGO, will be down in our place. I came to Hangzhou last September, to the Cloud Valley Center, and the office was officially opened in June this year, but there was no ceremony, because it was busy.
A few days ago, a lot of people were watching the GOAI World Artificial Intelligence Opening Competition, and dozens of people were busy with it. It's only a little bit of time in October. It's exactly 1024 today, and it's a very nice day for our programmer to do something like this.
It's an ex-summary. I'll take you 10 minutes to brief you on some of the things that are going on across the state of polarisation, including the Moot. Don't say too many big things, just my own little thoughts.
It is now recognized that AI development is in three stages. If you look at some IT or AI news, the probability is that over the last three years, and especially before last year, it was all LLM, and you've been working very hard on a variety of parameters. But do you guys feel like a little cooling this year? Where's the whole heat? There are two small examples that you can feel.
One of the original Teambit founders, Zi Joon, went to start his own business and became personal Agent, a personal AI agent, should be known as Today AI, and now the whole valuation is $1 billion. Turns out one of Ali's classmates, Lin Joon-hyun, is going to AI agent at the enterprise level, with an overall valuation of $5 billion.
So, you can probably feel that AI is entering a full range of fields, entering this phase of AI agent. Why do many people call this year "Yen AI agent"? I think it has something to do with this.
Of course, for the future, it's not too much to speculate about. AI is really going too fast. But in my view, as more and more AI agent and Agent appear, there is probably an AI operating system, or an operating system dedicated to AI agent.
The hot spot right now is, how is AI used in a thousand businesses. So, on the basis of this judgement, Polar Technologies has also upgraded its vision mission, adding a very critical phrase, "Assisting Enterprise AI to accelerate landing."
That is to say, in the coming years, our main focus will be to help businesses use AI internally, whether workflow or otherwise, to see if they can find the second growth curve through AI, achieve efficiency gains and improve.
Of course, this is not a new concept, and many businesses are saying it. Including our company, we've done a lot of work on this for the past year or two. For example, when you're doing the AI landings, you're putting together more representative cases across the industry. The brochure will be available in the near future if you are concerned about the timing of the event.
We also made an AI engineer knowledge base, which includes applications of LLM, AI agent. In this way, it is hoped that the whole industry will be made aware of what is happening and what cases are truly relevant.
If you're an AI engineer, how to use LLM, how to use these tools can have a reference. Because the whole state of polarism is basically these AI engineers. If you're interested, you can scan the two-dimensional code.
For business services, the Optimus Time enterprise version is a major force. Like China's movement, where hundreds of thousands of people are trained, the next four years will be hard-working. We're doing it for 380,000 people, from the chairman of the board to the AI training for the grass-roots staff.
But since the beginning of this year, we have found that there has been a major change in the industry. You might not just say that you're here to give me a few days of training, but rather that you can't do some AI agent, some Skill with our engineers, our team.
So, based on these judgments, we have also upgraded the services provided to businesses by AI agent, Agent, into AI Talent Management and Development, AI Training and Learning, AI Enterprise Solutions, using almost eight AI agent, to make the whole change.
As a result, there is a greater preference for AI and AI agent in terms of skills upgrading and team organization efficiency within enterprises.
What I have just said is mainly some of the business we have done in the past. But in my opinion, that's definitely not enough.
InfoQ provides information services, and kiosks provide knowledge services. I used to give an example: the so-called information service, like you're in a big square, telling people with loud horns, AI is important, AI agent is important, listen.
And more of all, I know it's important, and I'll teach you how to use it. But you've learned, you've used it, you've got a landing scene. And that's what a model workshop is all about: it's really an AI landing service platform.
If one sentence is introduced, it is to translate some of the AI ideas of the enterprise into real business results.
Of course, the challenge is to change the corporate vision or upgrade a business.
So back to back, what are we trying to solve? Or split it in two.
This year coincides with the ninth anniversary of Optimus Time, which has served almost 4 million individual users and over 4,000 businesses over the past nine years. In the last year or two, I have also visited many businesses to deal with them and communicate with them. And in the process, they found a very obvious pain: I have a budget, I have time, I have a scene, but I don't know where to find a solid AI delivery team to help me solve the problem.
What worries them most is that I gave you time and money without getting a landing.
I'm an entrepreneur myself. The biggest cost for business is actually time. You've made the wrong decision, but your friends may already be in the front, chasing after them, and it's gonna be harder.
For many of the AI delivery teams, including many of the FDEs, there's actually a pain: I know AI, I know technology, I have time, I'm willing to serve you. But what I'm worried about is, do you think you can get the budget you promised?
So, even though the Internet has developed over the years, we have spent a lot of effort addressing the issue of trust, but it still exists. I was just wondering, is it possible to solve this problem through model workshops, by way of AI?
One might think that an enterprise has an AI scenario and a team that wants to serve it, but it may not be clear to both sides how much time and budget it takes to complete.
If one day, when an enterprise needs publish in a model workshop, AI's supply and demand brain can tell it more easily how much time and budget it will take to complete it; can it also recommend a relatively reliable team on the platform and tell it that these teams, or these people, have a good reputation to deliver in the past, and have a good team strength?
For resident AI teams or super-individuals, they can see on the platform which businesses are willing to spend as much as they want, solve what problems, and choose which ones have relatively better credibility.
It is not only the intelligent matching of needs, but also the measurement of results when delivered. I think it's going to take a lot of time to do what we're going to do to connect the two parties through AI.
Of course, many would ask that this business pattern is not scarce, that it has existed for a long time in the Internet era and that many people are doing it. Why is it that there is a problem with the extremist state and that it thinks it can do it well?
I also listed a few reasons.
First of all, next year is the 20th anniversary of InfoQ entering China, and we think we know about technology, about AI.
And as we mentioned earlier, if we add the TGO event, the entire time of the minibus, we've served over 4 million individual users and 4,000 businesses. That is why we think that we have a relative appreciation for talent and business.
You know AI, you know talent, you know business, and you accumulate these better resources. So, we think it's our responsibility to come forward, to play such a role, to perform such a mission, to really get AI into a business.
But is this going to be done only by the family? Neither do I, especially in this particularly ecological era. So we've also begun to select some ecological partners to work together for the business.
A few simple lines here.
The first is a deep order, the founder of which is Dr. Yeh Jun, former President of the United States of America. Today, its co-founder and head of our TGO Zhangzhou branch, Joo-Fang, Boon Mountain, came to the scene and will share it later.
The second is Dr. Li Kefeng of Digital Technology, a former CTO retailer in Kyoto. And Suh Abo, who is supposed to be a professor, is the number of stories behind him; they include the Savannah's trick data and Zhang's Corinth.
These eco-partners are listed with a common feature: they're all our TGO membership companies.
That's why we're here. We're taking advantage of the ecological resources accumulated over the past 20 years by Polar Technologies to get the company AI down.
Of course, there are many small partners who say that we want to be involved, and how should we be involved? Very simple, scans involved.
Just kidding. There are two kinds of situations: if you have the capability to land, you want to find some scenes here that can be our AI ecological partner, like the deep order, digital technology, storytelling.
If you're an enterprise, you have a very good AI scene, you want to find a more reliable solution or an AI delivery team, you can also scan it. This is a typical bilateral platform.
We want to do enough with more eco-partners to get the company AI down there.
Finally, since today is 1024, wish all programmers happy holidays again! Of course, it's called AI Builder, Happy Holiday! And I want you to learn to be happy today.
Thank you again for coming. Thank you.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.