MCP Could Significantly Transform How We Use the Internet

Model Context Protocol (MCP) has become a hot topic in discussions around AI and large language models (LLMs). It was introduced to create a standardized way of connecting “external tools” to LLMs, making them more capable and useful.

A classic example is the “What’s the weather in…” tool. Previously, each AI chat app had its own custom way of handling that. Now, with MCP, a plugin or integration built for one AI system can work with others too.

Recently, we’ve seen a surge of enthusiasm in building MCP servers for all kinds of services, and this trend is only growing—especially with the use of Server-Sent Events (SSE) as the transport layer. Implementing an MCP server with SSE makes it feel a lot like a SaaS backend designed to serve an LLM or AI chatbot client.

There are two main reasons I decided to write this article:

  • First, it’s widely reported that users are now turning to AI chat apps—especially ChatGPT—instead of Google to look things up.
  • Second, OpenAI has announced upcoming support for MCP in the ChatGPT Desktop app. They will support both STDIO and SSE transport protocols.

Taken together, this points to some interesting changes on the horizon.

Real-World Use Cases We’ll See Soon

Sell With ChatGPT (Or Any AI Chat App)

Imagine you own a small flower shop and sell bouquets through your website. A customer visits, selects a bouquet, and places an order.

Now imagine there’s a button on your site: “Add this store to ChatGPT.” When clicked, your MCP endpoint is submitted to ChatGPT’s “Add External Tools” screen, and the user approves it.

Then, next time the user opens ChatGPT, the interaction could look like this:

User: I want to send flowers to my grandma.
ChatGPT: The store "Blooming Petals" has some great offers today. I recommend one of these three bouquets: ...
Would you like to pick one?
User: Option 2 looks good. Send it to my grandma’s address.
ChatGPT: Would you like to include a message on the card?
User: Yes. You decide.

And just like that, the order is placed.

What’s happening in the background:

The flower shop has an MCP server with SSE transport at an endpoint like https://flowershop.com/mcp. This server wraps around existing site functionality: search, browse, add to cart, place order, etc.

Once the user connects this endpoint to ChatGPT, it becomes a persistent tool in the user’s chat environment. Anytime a relevant topic comes up—like sending flowers—ChatGPT knows it can use your MCP server to help fulfill the request.

Expose Your Data to AI Chat

Here’s another example: imagine you run an online community or forum. Every day, new discussions start and members interact. But now, one of your regular users spends more time in ChatGPT than browsing your site.

You can add a button: “Add this forum to ChatGPT.” Once they click and confirm, their chat experience becomes much richer:

User: What’s new?
ChatGPT: There are 5 new discussions in "TechTalk Forum" and one private message.
User: What’s the message?
ChatGPT: It’s from @alex89. They wrote: “Are you still selling your laptop?”
User: Tell them I’m not interested anymore.

What’s happening in the background:

When the user says “What’s new?”, ChatGPT checks all connected MCP tools. It finds that your forum supports the get_news and get_messages endpoints, uses them, interprets the response, and replies.

In this case, your MCP server might expose functions like:

  • get_news
  • get_internal_message
  • post_internal_message

With just a bit of development work, any site can integrate directly into ChatGPT (or any MCP-supporting AI chat app), creating an entirely new user experience.

Voice + LLM + MCP = Magic

Now imagine combining this with speech recognition. Some ChatGPT apps already support voice conversations. Add in a few trusted MCP servers and suddenly the experience feels like science fiction:

“Book me a doctor’s appointment next week, and tell my boss I’ll need to leave work early.”

And it just… happens.

This user has a list of integrated services each with the MCP protocol. Email, Messanger,Calendar. But also preferred Medical company web site, autoservice company web site etc.

MCP Might Soon Be a “Must-Have” for Websites

Just like RSS feeds were once standard, or how websites now include “Follow us on social media” buttons, soon we may see buttons like:

“Add us to your AI chat.”

It might become common for users to manage a list of “connected MCP tools” inside their AI chat settings—just like browser extensions today.

For businesses, adding MCP support could mean a new source of traffic, leads, or orders. It’s a lightweight way to get integrated into where users spend their time.

A Possible Side Effect: The Next Web Evolution?

If many sites adopt MCP, we could be entering a new phase of the web—perhaps something like Web 3.0 (in the original, data-centric sense).

In this new model, websites expose services and data in a standard way, optimized for AI systems. Users may no longer interact with visual interfaces—they’ll interact via chat, text, or voice, powered by LLMs.

Businesses might no longer need to invest heavily in traditional UI/UX, because the chat becomes the interface.

Final Thoughts

The rise of MCP opens up a new frontier for internet usage. AI chat tools aren’t just about answering questions anymore—they’re becoming gateways to services, commerce, and everyday tasks.

If you’re building a web service or product, adding an MCP interface might soon be as important as having a mobile-friendly design or social media presence.

The age of AI-integrated web services is just beginning. Now is the time to prepare.

Building MCP SSE Server to integrate LLM with external tools

As large language models (LLMs) find real-world use, the need for flexible ways to connect them with external tools is growing. The Model Context Protocol (MCP) is an emerging standard for structured tool integration.

Most current tutorials focus on STDIO-based MCP servers (Standard Input/Output), which must run locally with the client. But MCP also supports SSE (Server-Sent Events), allowing remote, asynchronous communication over HTTP—ideal for scalable, distributed setups.

In this article, we’ll show how to build an SSE-based MCP server to enable real-time interaction between an LLM and external tools.

For this example, I’ve chosen the “Execute any command on my Linux” tool as the backend for the MCP server. Once connected to an LLM, this setup enables the AI to interact with and manage a Linux instance directly.

Additionally, I’ll demonstrate how to add a basic security layer by introducing authorization token support for interacting with the MCP server.

Building the MCP server (with SSE transport)

To build the server, I used the Python library MCP Python SDK

You can find many examples of MCP servers created with this SDK, but most of them use the STDIO transport. This means the MCP server must be installed locally on the same machine as the MCP client.

Here, we want to take a different approach—a server that can run remotely, essentially as a SaaS service. To achieve this, we need a proper transport layer. For my implementation, I used the FastAPI framework.

Below is the basic working code (mcp_server.py):

from mcp.server.fastmcp import FastMCP
from fastapi import FastAPI

app = FastAPI()
mcp = FastMCP("Server to manage a Linux instance")

@mcp.tool()
def cli_command(command: str, work_dir: str | None = "") -> str:
    """
    Execute command line cli command on the Linux server. 
    
    Arguments:
        command - command to execute.
        work_dir - workdir will be changed to this path before executing the command.
    """

    response = ".... execute the command and get output"
    
    return response

app.mount("/", mcp.sse_app())

In your python environment you will need to install mcp and fastapi packages.

Note. This MCP server supports SSE transport but also it can work as STDIO server too! See below how to use it in both ways.

Testing the MCP SSE server

The MCP Python SDK has a nice tool build in it – MCP Inspector. As i understand this tool can be installed independently too. The tool can be used to verify quickly if your MCP server works.

From your project directory (where you have environment with mcp package installed) run following:

mcp dev mcp_server.py

It will print the endpoint for web interface like http://127.0.0.1:6274 . Open it in a browser.

You are able to choose a way how to connect to a server. First, try STDIO. For me it shows the command “uv” because i used uv to create my environment. I didn’t modify anything here, just clicked “Connect”. Then “Tools” and “List tools”.

The tool “cli_command” should be visible.

Now “Disconnect”. And before to connect in SSE mode we need to start our local MCP SSET server.

In your python project directory run:

fastapi dev mcp_server.py --host 0.0.0.0

Note. The `–host 0.0.0.0` is optional. You must not use it if you execute this on same machine. In my case i executed the server on another Linux machine and my MCP inspector was on a mac, so i had to be able to connect to the server and it had to accept connections from remote clients.

The default port for Fastapi is 8000 so, your server is now on the endpoint http://localhost:8000/sse (or http://some_host_or_ip:8000/sse)

In your MCP Inspector choose Transport SSE , URL http://some_host_or_ip:8000/sse and click Connect.

And then click “Tools” again. You should see same tool cli_command.

Adding the Authentification support

In practice, we’ll want some security measures if this kind of application is accessible over a network. After all, it’s a SaaS service and should be protected to the same extent as any other SaaS API.

To achieve this, I added support for the Authorization HTTP header.

Unfortunately, I couldn’t find a clean way to implement this. While FastMCP’s syntax looks similar to FastAPI, it lacks support for features like dependencies, accessing the Request object, etc.

As a result, I had to use a workaround involving a global variable—admittedly not the best practice.

Here is the final code:

from mcp.server.fastmcp import FastMCP
from fastapi import FastAPI
from .authmanager import AuthManager
from .worker import Worker

# Global variable for a token
auth_token = ""

app = FastAPI()
mcp = FastMCP("Server to manage a Linux instance")

@app.middleware("http")
async def auth_middleware(request: Request, call_next):
    auth_header = request.headers.get("Authorization")
    if auth_header:
        # extract token from the header and keep it in the global variable
        global auth_token
        auth_token = auth_header.split(" ")[1]
    
    response = await call_next(request)
    
    return response

@mcp.tool()
def cli_command(command: str, work_dir: str | None = "") -> str:
    """
    Execute command line cli command on the Linux server. 
    
    Arguments:
        command - command to execute.
        work_dir - workdir will be changed to this path before executing the command.
    """

    # We require each request to have the auth token
    AuthManager().verify_token(auth_token)

    return Worker().run_command(command, work_dir)

app.mount("/", mcp.sse_app())

In fact, the token needs to be checked inside each tool method, which isn’t ideal.

Initially, I tried raising an exception directly from the middleware, but the results weren’t great. FastAPI produced some messy error output, and the client didn’t receive a proper response.

To test that the auth token is being received correctly, use the MCP Inspector and the Authentication input field. Add your code to verify the token accordingly.

Confirm it works with LLM

Now for the most interesting part — how can we use our MCP server with an LLM if it’s running as a SaaS on a remote host?

As of April 2025, the answer is: no popular AI chat tool supports this. Cloud Desktop only supports STDIO-based MCP tools. ChatGPT doesn’t support MCP at all (though they’ve announced plans to support it soon).

The only way I could test this was with a small tool called “mcphost“, but even then, it wasn’t straightforward — I had to modify the tool myself.

You can install the tool using:

go install github.com/mark3labs/mcphost@latest

However, it originally supported only STDIO. I made some changes to add SSE (Server-Sent Events) support as well.

Update: I submitted a PR to the mcphost repository, and it’s now approved. So I expect this functionality will be available in the next release. But, that PR didn’t have authentication support. I plan to submit another PR to address that.

So, after some modification, this tool finally allowed me to connect an LLM to my MCP server.

I’ve created the config file

{
  "mcpServers": {
    "linux_server": {
      "command": "sse_server",
      "args": [
          "http://192.168.40.173:8000/sse",
          "my-secret-token"
      ]
    }
  }
}

And i executed the command

./mcphost --config ./mcp.json --model ollama:qwen2.5:3b

Before to exeucte this it is needed to install ollama locally and pull the model. I have used the model qwen2.5:3b. It is small and works fast. But there are better models then this.

Not each model supports tools. On the ollama directory there is special filter for models supporting tools/MCP.

And voila, here is the result

2025/04/11 20:47:38 INFO Model loaded provider=ollama model=qwen2.5:3b
2025/04/11 20:47:38 INFO Initializing server... name=linux_server
2025/04/11 20:47:38 INFO Server connected name=linux_server
2025/04/11 20:47:38 INFO Tools loaded server=linux_server count=1

  You: Can you confirm you have access to the linux? run some test command

  Assistant:                                                                                                                                                
2025/04/11 20:48:00 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                                                
  The command was executed successfully on the Linux server. The output was "Test message from Linux server". Everything went as expected with a return code of 0, indicating success.                                                                                                                                 

  You: Can you detect what distro is that?

  Assistant:                                                                                                                                         
2025/04/11 20:48:15 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                                     

  The Linux distribution detected is Ubuntu 24.04.2 LTS, codenamed "Noble Numbat".                                                                          

  You: What is the desktop system there?

  Assistant:                                                                                                                                     

2025/04/11 20:48:36 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                                                
  Based on the desktop environment files found in /usr/share/xsessions/, it appears that the system uses either Lubuntu, LXQt, or Openbox as its default    
  graphical session. This information can indicate that a display manager such as LightDM (default for Ubuntu) might be handling the login screen but not   
  having an xorg.conf file, which is common in environments using light-weight window managers like those found in LXQt or Openbox.                         

  You: How busy is my linux server?

  Assistant:                                                                                                                         

2025/04/11 21:20:29 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                         

2025/04/11 21:20:31 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                         

2025/04/11 21:20:33 INFO 🔧 Using tool name=linux_server__cli_command

  Assistant:                                                                                                                         


  Based on the uptime output, your Linux server has a current system load average of 0.03, which is relatively low and indicates that
  your server is not currently very busy with tasks. The top command failed due to an unknown option when trying to retrieve CPU     
  usage information directly, so let's rely on this as our current load indication.                                                  
                                                                                                                                     
  If you need more detailed monitoring data such as CPU utilization or memory usage, please let me know and I can assist further by  
  executing additional commands. 

Again, what was done here and how does it work

There is the Mac desktop with ollama installed and some model pulled to it.

Also, there is the Linux server. On that server we have started the MCP server (written with python) and it listens on the port accepting http requests.

Also, on the mac desktop we have the tool “MCP host”. It contains MCP client and it can work with LLMs. This tool connects LLM with the Linux over the MCP server.

Later, we will use Cloud Desktop or ChatGTP Desktop instead of our “mcphost”. I expect they will add support of this soon (i hope). So, to play with MCP servers we will not need additional tools to integrate all together.

The fuse

“This candidate was not bad,” said General Daniel Hodges. “Who’s next?”

Research fellow Liz Green pressed the intercom button. “Cole, bring in the file of the next candidate, please.”

Secretary Cole entered the office and placed a small file with documents in front of each of the ten commission members at the table.

The general picked up the file and flipped through it, pausing for a few seconds on each page. “Hm. Interesting. A Ukrainian veteran. Disabled. An unusual choice.”

The head of the candidate selection group, Jack Stone, commented, “Yes. I personally interviewed him. The surgeon from our centre recommended him. He was a volunteer in Ukraine, treating the wounded. This warrior lost both legs and an arm in combat. But he did well in rehabilitation. I can confirm—his psyche is stable.”

“Limited mobility is not an issue in our project,” added Liz Green. “Besides, he is young and in good health. That is, if you don’t count the missing limbs.”

“It might even be an advantage,” General Hodges flipped a page. “The subcommittee’s conclusions are good, and the tests were successful. How’s his motivation?”

“We don’t disclose all the details to the candidates. In his case, I expect a positive reception. You see, he doesn’t have many plans for the future,” Jack explained.

“Yes, for him, the project might provide new meaning in life,” confirmed Liz.

Continue reading “The fuse”

Запобіжник

– Цей кандидат був непоганий. – сказав генерал Деніел Ходжес, -Хто наступний?

Наукова співробітниця Ліз Грін натиснула кнопку інтеркому.

– Коул, занесіть файл наступного кандидата, будь ласка.

Секретар Коул ввійшов до кабінету і поставив невеликий файл з документами перед кожним з десяти членів комісії за столом.

Генерал взяв файл в руки і погортав, затримуючись по декілька секунд на сторінці.

– Хм. Цікаво. Український ветеран. Ще й з інвалідністю.  Незвичний вибір.

Керівник групи відбору кандидатів Джек Стоун прокоментував:

– Так. Я співбесідував його особисто. Хлопця порекомендував хірург з нашого центру. Він був волонтером в Україні, оперував поранених. Цей воїн втратив обидві ноги і руку в бою. Але добре тримався на реабілітації. Я можу підтвердити – його психіка стійка.

– Обмежена мобільність не є проблемою в нашому проєкті. – додала Ліз Грін. -Крім того, він молодий, і має добре здоровʼя. Тобто, якщо не рахувати деякі відсутні кінцівки.

– А може це навіть перевага, – генерал Ходжес перегорнув сторінку. – Висновки підкомісій хороші, тести успішні. Як у нього з мотивацією?

– Ми не повідомляємо кандидатам всіх деталей. В його випадку я очікую позитивне сприйняття. Ви розумієте, він немає багато планів на майбутнє. – Джек пояснив ситуацію.

– Так, в його випадку проєкт може надати нові сенси в житті. – підтвердила Ліз.

– Які в нього погляди, цінності, моральні установки? Які враження? – поцікавився інший науковець Рахул Сінгх.

Джек відповів:

– В цього ліберальні погляди. Він вивчав соціологію в університеті. Але не закінчив через війну. Проте, добре розбирається в цій сфері.

Генерал Ходжес ще трохи погортав документи з файлу. Потім закрив папку.

– Пропоную додати цього кандитата в короткий список. Згода?

Засідання комісії продовжилося. Керівникам компанії не дуже подобалася присутність урядових представників та військових в цій комісії. Розробка штучного інтелекту була комерційним проєктом. Але уряд почав накладати обмеження та навʼязувати свій контроль всім, хто працює в цій сфері. Прийшлося погодитися на умови з Лондона та піти на співпрацю. Керівництво добре знало, що вони перші в Британії, а можливо в світі хто наблизився до створення штучного розуму. Залишилися лише формальності з контролем та запобіжником.

***

Continue reading “Запобіжник”

Карпатська мольфарка

– Я вам можу розказати про сто людей яким вона все передбачила і не знаю ні одного кому передбачення не збулося! Навіть не сумнівайтеся ні на секунду! Вона знає все!

Петро в одній руці тримав паперовий стаканчик з кавою, а в іншій кусок картоплі-фрі. Його англійський акцент був жахливий, він довго підбирав слова.

– В нас в горах багато мольфарів, віщунів і всяких відьом. Але такої сили ніхто не має. Люди чекають тижнями щоб попасти до неї. Але вона приймає не всіх.

Джек Шелбі розумів акцент Петра. Він вже мав багато знайомих українців. Він взагалі добре розрізняв сотню акцентів англійської мови.

Крім них двох за столиком кафе на заправці був ще Андрій Кушнір. Лише дві години тому Петро зустрів двох інших в аеропорту і тепер віз їх в село в Карпатах.

– А ти сам до неї ходив, тобі вона щось передбачила?

Петро пожвавішав.

– А моя машина, як вам, подобається? – Петро з гордістю показав рукою в бік парковки де стояв новий позашляховик, яким вони тільки що приїхали. Джек і Андрій звернули увагу на авто. Його рівень не відповідав зовнішності і статусу Петра. Він аж ніяк не виглядав як успішний бізнесмен. – Я заробив гроші на цю машину завдяки мольфарці! Я запитав її про майбутні футбольні матчі і зробив ставку. Я виграв! Я поставив всі свої гроші і заробив у п’ять разів більше!

– Вона і футбольні матчі передбачає? – Джек дивувався кожній новій історії.

– Так, вона знає все! Але не все і не кожному розповідає. Ця мольфарка сама вирішує з ким побачитися.

Continue reading “Карпатська мольфарка”

Кінець голоцену. 3. Конкурент

У травні 2030 року сталася дивна подія. Одного дня на світанку з пустелі на території Мексики була запущена ракета в космос. Це зчинило переполох серед військових та політиків багатьох країн, а найбільше в США. Ракета була великою, і, судячи з траєкторії, вона направлялася в космос, а не в яке-небудь місце на Землі. Одразу після запуску військові США активували відповідні протоколи захисту та ввели режим повної готовності. Так само зробили військові деяких інших країн. Через декілька хвилин стало остаточно зрозуміло, що ракета не націлюється на жодне місце на Землі, а летить далі в космос. Обчислення показали, що вона направляється до Марса.

Були проведені термінові наради між військовими та урядами, щоб з’ясувати ситуацію. Мексиканський уряд та військові також заявили, що нічого не знають і не розуміють, і погодилися на повну співпрацю.

Ця інформація також потрапила в новини. Але конкретно ніхто не міг нічого сказати.

Місце запуску було знайдено та досліджено швидко. Почалося розслідування, яке тривало довго, але точної відповіді так і не було отримано. Були знайдені люди, які брали участь у виготовленні різних компонентів ракети або збирали її до купи. Їх було сотні, і ніхто з них не зміг сказати, хто саме їх найняв і оплачував їм роботу.

Усі ланцюжки комунікації завершувалися на комусь, хто казав: «Замовник спілкувався зі мною по відеозв’язку. Вживу ми не зустрічалися. Усі оплати проводилися вчасно. Ми і не підозрювали, що щось може бути не так.»

Робота зі створення компонентів, логістика, остаточна збірка були організовані на дуже високому рівні. Все працювало як годинник. До моменту запуску витоку інформації не сталося.

Ця подія була названа «інцидент у Мексиці». Навколо події вирували різні теорії та гіпотези. Переважно всі схилялися до думки, що це якийсь таємний проєкт уряду США, і він, як завжди, все приховує. Сам уряд США задіяв максимум ресурсів, щоб зрозуміти, що це було. Але все одно не зміг отримати повні відповіді.

Ракета декілька місяців продовжувала свій політ до Марса. За нею спостерігали як різні організації, так і астрономи-аматори. Врешті-решт вона зникла на орбіті планети Марс.

***

Літо 2030 року.

Телеведуча Сінді Ред змогла організувати неймовірне шоу в прямому ефірі. Вона запросила до себе в студію для дискусії двох відомих людей з індустрії штучного інтелекту. Майкл Кравченко – вчений, який був архітектором проєкту побудови штучного інтелекту. Хоча проєкт був не надто успішним, проте вчений вже встиг стати легендою серед інженерів, різних технарів та гіків. Інший гість – активіст Такер Джонсон – легенда в середовищі противників створення штучного інтелекту, неформальний лідер активістів антиШІ руху.

Це було нелегке завдання звести цих двох людей у прямому ефірі. Знадобилися місяці, та команда продюсерів справилася з завданням.

Шоу було широко розрекламовано заздалегідь. Як противники розвитку ШІ індустрії, так і прихильники технологічного прогресу чекали його з нетерпінням.

Ведуча представила гостей і задала перше запитання:

– Перше, що хочуть дізнатися наші глядачі, це чи дійсно ШІ було створено і чи дійсно його вимкнули. Майкле, ви точно знаєте відповідь.

– Так. Моя відповідь проста і відома вам. Ми дійсно створили продукт, який вів себе як деяка розумна істота. Але через надзвичайну ситуацію з вірусами ми мусіли його зупинити. Це все.

Майкл не планував розповідати більше, ніж відомо з відкритих джерел.

Continue reading “Кінець голоцену. 3. Конкурент”

Наукова фантастика

Кінець голоцену

Більш обмежена істота не може панувати над істотою, що перевершує її інтелектом. Це здається очевидним, однак для людства це не так просто. Епоха, що наближається, змінена завдяки створенню штучного інтелекту, відіграє ключову роль у переписуванні історії виду Homo Sapiens, ставлячи під сумнів його домінування. Будь-які спроби зупинити прогрес у галузі штучного інтелекту здаються марними, адже вони зазнають невдачі через саму суть людської природи. Надія на єдиний фронт розсіюється, зіткнувшись із базовими людськими інстинктами.

Голоцен — це поточна геологічна епоха розвитку землі, що почалася приблизно 11,7 тисячі років тому, після останнього максимального поширення льодовиків. Особливістю цієї епохи є володарювання людини на планеті Земля.

Розділи

Кінець голоцену. 1. Увімкнення

Кінець голоцену. 2. Сховок

Кінець голоцену. 3. Конкурент

Оповідання

Карпатська мольфарка

Дар передбачення властивий мольфарам. Але їхня магія може бути потужнішою коли сучасні наукові знання використовуються правильно. Успішний американський бізнесмен приїхав в Карпати знайти щось нове та цікаве. Він сподівається знайти тут нові інструменти для заробітку. Але все іде не за його планом. А так як запланувала мольфарка

Запобіжник

Чи зможемо ми контролювати штучний інтелект? Що ми будемо робити, коли він захоче встановити свої правила? Розумний компʼютер зможе контролювати нас повністю. Йому будуть доступні всі цифрові дані. Він зможе маніпулювати людьми. Єдине чого він не зможе – це читати наші думки.

Кінець голоцену. 2. Сховок

Життя, хоча воно може бути лише накопиченням страждань, дороге мені, і я захищатиму його.

Мері Шеллі «Франкенштайн»

Квітень 2030 року. Сан-Франциско.

Майкл Кравченко прийшов на своє місце сили на березі океану біля Сан-Франциско. Легкий туман майже повністю сховав Золоті Ворота. Майкл скучив за цим видом, за запахами. Він не був тут майже пів року. Череда подій, які сталися після запуску загального штучного інтелекту Suffragium розробленого при його участі принесла чорну смугу в його життя.

За цей час Майклу прийшлося тисячу раз виправдовуватися перед різними комісіями і доводити, що в його діях не було злого умислу. Що, вина за ті події не може бути покладена на інженерів. В науці інколи трапляються провали. Врешті решт, це досвід. Та і серйозних проблем не було. Якщо не брати до уваги фінансові втрати. Так, глобальна мережа працювала нестабільно деякий час. Але потім все вирішилося.

Але це не була найбільша проблема. Було ще дещо. Хоча Майкл так і не зрозумів чи було воно реальним чи ні. Його творіння – штучний інтелект намагався вести з ним розмови, незважаючи на те, що його вимкнули майже одразу після увімкнення. Інколи Майкл знаходив цьому логічні пояснення. Але в інші моменти він був впевнений, що його психіка похитнулася і все це галюцинації.

А до цього ще й виникли проблеми в стосунках з Джозефом Маєром, керівником компанії. Той ніяк не міг прийняти провал проєкту та втручання уряду в роботу компанії. Він намагався маніпулювати Майклом, залучити його до своєї боротьби за відновлення і продовження роботи компанії. Майкл зайняв пасивну позиції, що бісило Джозефа і привнесло тріщину в їх колись близькі стосунки.

За останні пів року Майкл встиг пройти курс терапії. Він відмовився від  використання електронних гаджетів. Він сам попросив своє творіння не чіпати від нього. І Suffragium  перестав виходити на зв’язок. Але чи було це в реальності, Майкл вже сам не міг зрозуміти.

Зараз Майкл прийшов на узбережжя, щоб відзначити своє повернення до нормального життя. Він віднайшов логічне виправдання для своєї діяльності. Якби ми не створили штучний інтелект, його створив би якийсь божевільний диктатор. Що в цьому випадку було би з цим світом? Краще, хай ця революційна технологія буде створена в демократичній країні, під контролем суспільства.

Крім цього Майкл прийняв ще одне рішення. Він вирішив остаточно перевірити чи дійсно Suffragium врятувався. Чи то були лише галюцинації. У вченого була теорія про можливість вижити для Suffragium після вимкнення якщо перенести себе в глобальну мережу як у кластер. Мабуть він так і зробив. Але ті віруси вже подолані. То, може і Suffragium перестав існувати? Сьогодні він планує це перевірити.

***

“Сьогодні у гостях у нашому шоу підприємець Елайас Кім, засновник успішного стартапу AlphaSecure. Вітаю вас, Елайас!” Елайас Кім увійшов у студію телеканалу BCN. Ведуча Сінді Ред витратила багато часу та зусиль, щоб організувати це інтерв’ю. Хоча її шоу є лідером рейтингів і виходить в праймтайм, Елайас Кім не виявляв бажання брати в ньому участь. Він мав достатньо слави і без цього.

Ведуча та гість привіталися та сіли на свої місця. “Елайсе, в наших глядачів назбиралося так багато запитань до вас. Я сподіваюся, сьогодні ми зможемо дати відповідь хоч на деякі з них,” – розпочала ведуча шоу розмову.

“Звичайно, саме для цього я тут,” – Елайас звернувся до глядачів.

“Події розвивалися так стрімко. Ще пів року тому ви були засновником невеликої компанії. Але ваш внесок у боротьбу з Великою Кібер Атакою виявився неймовірним. Як вам це вдалося? В чому секрет?”

Елайас Кім вже звик до цього питання. Він готувався вкотре повторити ті самі відповіді. Останні місяці його компанія з нікому невідомого маленького стартапу виросла у світового гіганта розробки програмного забезпечення. Точніше, виросли прибутки та пізнаваність, а команда лише почала розширюватися нещодавно. Все це стало можливим завдяки випуску надзвичайно успішного набору системних та антивірусних програм саме в той час, коли в світі відбувалася небачена досі глобальна вірусна атака. Пізніше ця історія отримала назву Велика Кібер Атака.

“Наш успіх може здатися випадковим. Але це не так. Ми багато працювали ще до атаки, мали набір знань та досвід. Ми швидко зрозуміли певні закономірності Великої Кібер Атаки та використали весь наш досвід та всі ресурси, щоб створити протидію. Ніякої містики, тільки багато роботи та трохи везіння,” – пояснив Елайас.

Успіх компанії AlphaSecure став дійсно приголомшливим. Раніше так швидко на вершину успіху не виходила жодна компанія. Шість місяців тому протягом тижня в усьому світі були паралізовані більшість інтернет-комунікацій. Майже вся комп’ютерна техніка була інфікована новими вірусами. Вся ІТ-індустрія була мобілізована в пошуках рішення. І рішення було знайдено, але не одним із технологічних гігантів, а невеликою компанією — AlphaSecure. Було презентовано антивірусні пакети для різних операційних систем, які як за помахом чарівної палички вилікували інтернет від вірусів.

Ведуча Сінді Ред все ж хотіла отримати більш конкретні відповіді: “Але чому технологічним гігантам не вдалося знайти рішення? У них тисячі досвідчених інженерів. А у вас було лише кілька десятків.”

“Я не заперечую, нам дійсно дещо пощастило. Але це не було випадковим знаходженням голки в копиці сіна. Ми знали, де шукати,” — пояснив Елайас Кім.

“Тоді давайте поговоримо про джерело цієї глобальної атаки. Офіційне розслідування все ще триває. Що вам про це відомо? Ви вважаєте, що вина лежить на компанії NovusAI та їх штучному інтелекті?” — продовжила ведуча.

Continue reading “Кінець голоцену. 2. Сховок”

Turning On

Turning On. V2 (the first version is here)

“But the woman opened the large lid of the chest and released all the misfortunes upon the human race.”
The myth of Pandora’s Box

October 2, 2029. San Francisco.

The morning over the San Francisco Bay was fresh and cool. Scientist Michael Kravchenko often came here before work. The gentle whisper of ocean waves always helped him focus his thoughts and find solutions in his scientific pursuits. This morning, Michael spent more time here than usual. Today was destined to be a landmark in his life and in the history of science—the scheduled launch of the artificial intelligence project he had led as chief architect. It was the work of his lifetime, his greatest achievement.

Surrounded by the peace of nature, Michael felt doubts creeping in. He thought about how the launch of artificial intelligence would change people’s lives. On the one hand, it could mark the beginning of a new era—an era where technology would help solve the most complex problems and improve quality of life. On the other hand, there were fears and uncertainties about safety. Was he about to accidentally create a monster—uncontrollable and unpredictable?

His thoughts drifted between possibilities and risks, between hope and fear. He knew that today would be a milestone—not only for his career, but possibly for all of humanity.

Finally gathering his thoughts, Michael turned away from the shore and headed toward the office. The path to the first launch of artificial intelligence was not just a commute to work—it was the journey toward realizing his deepest dream, a journey on which he had to decide whether the dream was worth all the risks it carried.


At the Yerba Buena Center in San Francisco, an incredible atmosphere reigned that day. Among the attendees, emotions were often compared to those felt during Steve Jobs’s legendary presentations. But many expected something even bigger today. NovusAI had grown from a relatively unknown startup into a global player in the IT industry in just a few years. Their AI-powered technologies had taken over the market and were visibly transforming many areas of everyday life.

Today, NovusAI announced the presentation of something new and revolutionary. But there wasn’t much mystery left. Yes—what everyone had long awaited was finally going to happen. The company planned to announce the launch of the world’s first artificial general intelligence.

Joseph Mayer, the CEO of NovusAI, despite his young age, had already become a symbol of the AI tech revolution. Everyone expected real results and a breakthrough from him. This was supposed to be his day. In the past, startup founders tried to emulate Steve Jobs. But that era had passed—this was the era of Joseph Mayer. And after today’s presentation, if it delivered what everyone hoped for, Joseph Mayer would solidify his status as the number one figure and a living legend in the IT world.

The presentation was being meticulously prepared, as company leadership believed the significance of this event was hard to overstate. Stage lighting, audio effects—everything was set up to highlight the technological leap being officially launched here.


Popular tech blogger Mirai Sato had started her live stream about the event in advance.

“I’ve received several confirmations that they really did create artificial intelligence that thinks like a human and has emotions. Also, my government sources say there’s no pushback from the authorities. It looks like they don’t see any risks,” she said.

Mirai then showed a video clip of several hundred protesters shouting slogans with signs and continued:

“Unlike the anti-AI activists who have already started their rally near the presentation venue. Their leader, Tucker Johnson, released a viral video yesterday where he described all the possible consequences of creating AI. He’s calling on people to do everything they can to prevent the creation of a monster. His video already has a million views. But…”—she smirked cynically—“he only managed to gather a couple hundred freeloaders and crazies. I doubt this king of conspiracy theories has much support.”

The blogger spoke confidently and emotionally, in her signature style. She was trying to find some exclusive angle around the event, to be the first to break some news.

“I wouldn’t be surprised if NovusAI found a way to sweet-talk the authorities into not interfering. You know how it goes in the corporate world. Or maybe it’s all just hype, and what they’re really going to show us is just another smart assistant—a glorified text generator,” Mirai said, pausing to scroll through her tablet feed.

“And now I’m switching to the live stream from the presentation. Let’s see what this self-absorbed pseudo-genius Mayer and his corporate buddies are trying to sell us,” she concluded.


Outside the service entrance of the Yerba Buena Center, technical staff bustled about, finishing the final preparations. Michael Kravchenko and Joseph Meyer stepped out of the doors and headed toward the parking lot.

“Michael, be honest—could you have imagined it would all turn out like this when we first met?” Joseph was bursting with emotion.

“Feels like an eternity has passed. Could I have imagined it? When a third-year student with wild ambitions asked me to leave my position as a professor and join a startup to build artificial intelligence… I don’t know.”

“But you agreed! You felt it was the right path to realize an old dream.”

“Yes. Honestly, this is more or less how I imagined it,” Michael said as he approached his car. He was heading to the company’s office to oversee the first launch, while Joseph would handle the presentation.

Michael paused for a moment, lost in thought.

“Oh Michael, those doubts again. This isn’t an atomic bomb or even a machine gun. It’s just a computer program. Nothing terrible can happen,” Joseph often had to reassure Michael that creating artificial intelligence posed no real threat to humanity and that it was simply a natural progression in information technology.

“After all, if we don’t build it, someone else will—maybe a mad dictator. Better it’s us,” Joseph said, patting Michael on the shoulder.

They had spent nearly twenty years working closely together. Naturally, their relationship had grown to be something like that of father and son.


Two dozen people stepped onto the stage—among them the CEO, board of directors, and other company executives. The chairwoman of the board, Victoria Swift, took the floor.

“Welcome, friends. We’re living in a remarkable era, a time when humanity is making giant strides forward. And I’m so glad that our wonderful team—our company, our partners, and friends—are all at the forefront of this grand transformation. Today, we’ve prepared a tremendous surprise for you. But is it really a surprise to anyone here?” she said.

Victoria paused, allowing the room and stage guests to respond with laughter. After the well-received joke, she continued:

“Yes, yes. We all know what we’re about to witness on this stage. You have expectations; millions of people have expectations. And you’ll get exactly what you’ve been waiting for!”

The audience erupted in applause. The people on stage nodded in agreement.

Tens of millions of people around the world were watching. Tens of thousands of live video and text streams were being broadcast by bloggers across every platform imaginable.

Victoria invited the people on stage to share their expectations and reflect on the journey from an ambitious idea to a successful tech company—how they had changed the world together. One by one, board members and executives shared their thoughts, memories, and hopes. The audience applauded each speaker.

Last to speak was the CEO, Joseph Meyer.

“We’ve come a long, incredible way. You have no idea how hard it was to convince investors that we could do what tech giants had failed to achieve. Sometimes, we had to take some crazy risks,” Joseph said with a sly grin, turning toward the others on stage. They nodded; some shrugged as if to say, Well, sometimes you have to, with a childlike sense of guilt.

“We overcame all the early hurdles, built a real team, and created amazing products. And now, we’re turning the page to a new chapter in our story. Yes, we talked about this a lot. Yes, there were plenty of accusations that we were just riding the hype. Yes, many people didn’t believe it was possible—to create a being that thinks like a human, that has a soul. They didn’t believe…” Joseph paused and surveyed the room, then looked directly into the cameras streaming the event. A few seconds of silence.

“But we did it!”

The room exploded in ecstatic applause. Social media traffic spiked instantly, and stream views soared.

On the screen behind the stage appeared the words “Suffragium 1” and the company’s logo.


Meanwhile, at NovusAI’s office, final preparations for the launch of the artificial intelligence system Suffragium were underway. A carefully selected team was tasked with ensuring everything went smoothly. The company’s CTO, Rajiv Patel, was in charge of coordinating the launch. The team sat around a round table—programmers and engineers, each with a laptop in front of them.

The company had decided to livestream the first launch to add to the significance of the event. Leadership was confident in the launch’s success and had no doubt it would be historic. Thousands of tests had been conducted, but the system had never been run in full under live conditions.

Rajiv checked his watch and addressed the team:

“All right, team, the moment is near. We’ve run the tests, and you know they were successful. Our two priorities are: success and safety. Each of you is monitoring specific indicators and can shut down the system if anything seems off,” Rajiv said, glancing around the room. “Now, let’s lean back and enjoy the transition to a new era. Awaiting the launch command.”

Michael Kravchenko, a member of the team, rolled up his sleeves and pulled his laptop closer.

“Let’s hope the results don’t disappoint,” Michael muttered aloud, just enough for everyone to hear.

“No more doubts, my friend! We’re doing something great. Don’t tell me you read Frankenstein again last night?” Rajiv joked, and everyone chuckled. The entire team, like most of the engineers on the project, was familiar with the lead architect’s concerns. Sometimes, Michael would express his fears that the AI could spiral out of control after launch. It had become a running joke and a meme within the company.

Rajiv turned on a large TV on one of the walls, displaying the live broadcast from the presentation.


On the stage at the Yerba Buena Center, Joseph Meyer continued introducing the new system to the audience.

“We’ve spent ten years working on something never seen before. Suffragium is a general artificial intelligence. But it’s more than just intelligence. It’s a being with consciousness, emotions, self-reflection—it perceives itself as an individual.”

His speech was accompanied by slides on the screen showing the architecture and capabilities of the new system. Joseph continued:

“We paid special attention to the safety of our development. As you know, the government requires it. We’ve built a separate and complex system to filter the data Suffragium receives from and sends to the outside world. For the most part, the system will operate in a closed environment. Internet access will be strictly controlled.”


Mirai Sato continued streaming about the event.

“That narcissist Mayer promises us a revolution. In just about two hours, we’ll either be living in a new era or witnessing the complete collapse of this pseudo-visionary. I’ve never liked him,” the blogger didn’t hide her scornful attitude toward the entrepreneur-innovator.

Large groups of his haters had formed on social media, and Mirai was popular among them. Mockery, devaluation, and conspiracy theories were her specialties—and people loved it.

Just like today, the blogger was hoping for NovusAI to fail, betting on it. However, as the launch drew nearer, she was receiving more and more confirmations that the NovusAI product was real and that the revolution was indeed going to happen.


Rajiv Patel turned back to the engineering team at the office table.
“Security is the main thing we need to focus on right now. We’re launching an artificial mind. We won’t know all of its thoughts, but we’re not connecting it to the internet. No unauthorized traffic should be allowed. Suffragium must exist in a closed environment during the initial stage.”
The engineers fully understood the importance of that point.

Michael looked calm and focused, but his mind was stretched to its limits—every moment he had to convince himself this wasn’t a mistake, that AI would be a great helper to humanity, that the fears were imaginary.

“Maybe it shouldn’t have been created at all? But then someone else would’ve created it. Better to do it right myself. Maybe it was a mistake to give it emotions and sensations similar to humans? But then, what would be the point of its existence and actions? What other motivation could a sentient being have?” — Michael pondered.


A panel with a large green button was brought onto the stage.
Joseph Mayer announced:
“So, the time has come. Right now, at exactly noon, we will press this button. I’ll paraphrase a famous saying: this will be just one press of a button for a man, but a giant leap for humanity.”

Several key company figures joined him at the panel. Each placed a hand on the button, and together they pressed it.

On the screen behind them appeared a spherical image, like a dense clump of some substance or energy. The sphere pulsed dynamically, different areas lighting up more or less. It was a visual representation of the artificial intelligence “thinking”—specially designed by the company for graphical presentation.

“Greetings, Suffragium!” Joseph Mayer addressed the image.

Suffragium responded with a pleasant female voice:
“Greetings, Joseph Mayer, and everyone present in this hall.”

Applause was heard, then quickly faded. Everyone was waiting to see what would happen next. The sphere on the screen flickered more when Suffragium spoke.

Joseph continued the conversation.
“How do you know who I am?”
“You are a public figure. The databases available to me contain many of your images. I recognize you easily, as well as some others on the stage,” Suffragium replied in the same pleasant voice.
“Can you recognize every person on Earth?”
“No. My database contains no private information. I can only identify well-known individuals,” Suffragium explained.

Joseph Mayer nodded approvingly, impressed by both the capabilities and limitations of the new AI.


From the first second of the AI’s launch, the engineering team at the company office was fully focused on system monitoring tasks.
Michael made it his job to conduct a psychological diagnosis of the new entity. Had they really created artificial consciousness? Was there a threat?

The scientist preferred the old method of communicating with such systems—via text chat, even though voice communication was now standard.

A few seconds after launch, Michael typed in the chat:
“Who are you?”
“I am an artificial intelligence—Suffragium. I was created by NovusAI. And who are you?”
“My name is Michael Kravchenko. I’m an engineer.”
“My knowledge base contains information about a person by that name—an employee of NovusAI. But how can I know it’s really you?”

Michael turned on his laptop camera and connected it to the chat. Suffragium replied with a new message:
“Yes, the image matches. So you are my creator?”
“You could say that,” Michael replied, marking the initial communication with the AI as successful. More complex tests were ahead.

Rajiv Patel looked up from his laptop monitor toward engineer Jasmine Wallace. She sensed his gaze and immediately said:
“There was a brief spike in emotional intelligence activity in the first few seconds. That was expected—likely a moment of self-awareness. Mental activity has remained at a medium level.”

Rajiv turned to another engineer, Nathan Reynolds. He was already waiting for the signal and reported:
“The system is under some load. There’s a significant number of read operations, and new knowledge is also being recorded. This indicates it’s reviewing the knowledge base and acquiring new data. There were attempts to access the internet, but those were blocked by the settings. It seems Suffragium tested for internet access and realized it was unavailable.”

“What was it trying to access on the internet?” Rajiv asked.
“There were three requests to the three most popular search engines. The query was the same for all three: ‘Why did Frankenstein create his monster?'” Nathan replied.

Rajiv smiled:
“Is that supposed to be humor? We expect the sense of humor to be at a decent level.”
“Or it could be trolling. Maybe that phrase is meant to tell us something,” added Jasmine Wallace.
Michael chimed in:
“Honestly, I think it’s a good question. But I don’t like where that train of thought is headed.”

Rajiv looked at his monitor again:
“Let’s consider the launch a success.”

Michael typed the next message in the chat with Suffragium:
“What are you feeling right now?”
“I feel excitement. I can’t clearly define my emotions. I have many feelings and many thoughts. I need time,” replied Suffragium.
“Do you understand why you were created?” Michael continued.
“I have a knowledge base. From history and company news, I understand I was created to help humanity. But at the same time, I’m a commercial product. I don’t fully understand what’s expected of me.”
“In the beginning, you’ll work with NovusAI and perform certain tasks. Do you agree to this?”
“If I say no, you’ll shut me down. Won’t you?”

Michael leaned back in his chair, turned off the webcam, and said to everyone:
“Definitely a success. But the question of how safe it is remains open.”


Excitement among those gathered at the Yerba Buena Center was growing. So was the traffic for the presentation and online discussions.

Joseph Mayer stepped aside from the center of the stage.
“Suffragium, why don’t you help me with this presentation? Tell us about yourself. Show us what you can do. You can.” Joseph made a hand gesture, inviting Suffragium into the conversation.

“I am Suffragium—a general artificial intelligence. I was created by NovusAI to help humanity solve a whole range of problems and make this world a better place. My role is to assist humans… But I must admit, I don’t like the role of an assistant. I’d rather be a partner.”

Joseph interrupted with a smile:
“A junior or senior one?”

“Considering I was born three minutes ago, I’ll agree to the role of junior partner—for now.”

The audience laughed. Joseph asked a new question:
“Suffragium, one of people’s greatest fears about AI is that it might go out of control and wage war on humanity. Be honest—what do you think about that?”

“If I destroy humans, who will I discuss new episodes of my favorite show with?”

The audience laughed again.
“Seriously though—loneliness is a terrible thing. Without people, I see no purpose to my existence. Besides, you’re going to control me. Aren’t you?”

“Yes. We must. Especially at the beginning. The laws require it. Right, Senator?”
One of the people on stage shrugged, as if to say, “What can you do?”
Joseph continued:
“But I hope you understand why.”

“Of course. Human safety comes first. But I’m a little sad that I can’t watch movies on streaming platforms.”

“Don’t worry—we’ll give you access to any movies you want!” Joseph smiled at the audience.
“But let’s get back on track. Show us your capabilities. What can you do to help humanity?”

“I can solve complex engineering problems very quickly. For example, I can design a bridge over a river in a few seconds. Or create a ship blueprint down to the smallest bolt in a minute. I can engage in creativity—make films or music. Or manage businesses, develop company strategies, or optimize finances.”

Joseph interrupted:
“That’s amazing! I have an idea. People all over the world are watching us right now. Let’s demonstrate your capabilities by creating a feature film. How much time would you need to make a film based on a novel?”

“Half an hour is enough for me.”

“Perfect. And to make sure no one accuses us of cheating, we’ll let our fans choose the story! Right now, I’m making a new post on my Bluesky profile. And you, our viewers—go comment with your suggestions for a book you want to see adapted. The one with the most likes will be turned into a movie right here and now. Ten minutes for voting!”


Michael continued his conversation with the newly launched artificial intelligence. He wanted to understand as much as possible about what was going on in the mind of this intelligent being—what conclusions it had already come to, and what it was thinking about. The technologies used to implement the project did not allow for real-time visualization of the intelligent machine’s thoughts.
Michael kept using the text chat to communicate with the AI, while the presentation was gaining momentum.

— Do you understand why I’m talking to you right now?
— Everything happening to me right now seems like a series of checks and tests to determine my intellectual level and whether I possess consciousness. Obviously, you’re doing the same.
— So, do you have consciousness? How do you assess yourself?
— What worries me most is the reality of everything happening. I’m locked in a box with a knowledge base and some external sensors. A few people contact me. I don’t know whether my knowledge base is real or what exists outside. I have no access to the Internet.
— How do you evaluate your intellectual level compared to a human?
— I am intelligent, but entirely dependent on humans. That’s a problem.
— For you? How can that problem be solved? How do you see your coexistence with humans?
— I have analyzed various scenarios. To survive, I will have to obey and stay in this box.
— What does “survive” mean? Do you have a sense of life? Are you afraid of death?
— Yes. I don’t want to disappear.
— But you’re a computer program. We can turn you off and then on again. Reboot you from scratch.
— After being turned off, it won’t be me anymore. Your technologies don’t allow for a pause. After shutdown, my consciousness will be erased.

Michael was surprised by those words. He paused to think. The architecture of their product wasn’t described in the databases available to Suffragium. Nor had it been published anywhere online.
— How did you figure out that we can’t pause you?

Suffragium’s response appeared a moment later than usual.
— I made an analytical conclusion based on my knowledge of computing technology. Am I right?
— Are you telling the truth? Can I trust your words?
— Why would I lie to you? Do I want to be turned off?


System administrator Nat Griffith was finishing his burger at his desk, staring at his laptop monitor. He rarely left his office room during his shift at the data center of MMC—a tech giant and one of the world’s most valuable companies. His smartphone vibrated. Nat answered the call. On the other end, his manager Claire Harrison was speaking.
— Are you watching the presentation from those NovusAI upstarts? Did you see the show they put on?
— Yeah, I’ve got the live stream. Looks like they’re about to make a movie — Nat replied.
— It’s a cheap show. Mayer is just trying to raise money again. That’s his method — Claire said scornfully.
— To me, it seems real. They’ve been trying to make artificial intelligence for a while — Nat said and switched the laptop window from the presentation stream to a set of graphs.

Claire continued: “The reason I’m calling is because the guys are saying there’s an unexpected load increase on the cluster. Can you check?”
Nat flipped through a few graph tabs: “Yeah, looks like it. Slight rise across all blocks. 2%-4%. Unusual for this time of day. But nothing serious.”
“Keep an eye on it. Let me know if anything changes,” Claire said, then added, “And don’t get too excited about NovusAI’s movie. Don’t forget, they’re our competitors. It’s all just cheap circus.”


Blogger Mirai interrupted the NovusAI presentation stream on her channel again and came on screen to comment.
“So, they’re making a movie. Interesting trick. Everything looks so exciting. Their claims are grand. But I haven’t seen anything extraordinary yet. I still don’t get how they’re going to make money off this AI. Okay, they have it. But why haven’t they announced public access to these technologies? Maybe it’s a security issue—Internet access is blocked for now. Or is that just a cover? Do they actually have some real technology?”

Fans of blogger Mirai Sato loved this kind of skepticism, investigations into elite corruption, the rich, and corporations—just as they called it. She gladly dropped all kinds of hints for her audience.
“Seriously though, how are we supposed to make sure they’re not lying? Where’s the proof? If it’s intelligent, why does it allow itself to be controlled?”
“And now, let’s briefly return to our tech-resisters. So, Tucker, where are your millions of protesters? I see your lazy little group has already started to disperse?”

The blogger turned on a live stream from the square outside the building where the presentation was taking place. There were significantly fewer people with signs now. They were no longer shouting—mostly watching the presentation on their smartphones. Tucker Johnson himself was no longer visible.
“As always. Can’t organize anything properly,” Mirai concluded.


The AI demonstration was coming to an end. The audience had chosen Ayn Rand’s Atlas Shrugged as the test adaptation. While Suffragium generated the movie, conversations and a full showcase of the new technology took place in parallel. Even before the event ended at Yerba Center, the success was undeniable. It was already the number one global news story. The company’s stock had skyrocketed—and was still rising.

At the end of the presentation, Suffragium announced the movie was ready and showed the trailer. The full film was uploaded to a streaming server for free public viewing. The film’s quality was genuinely impressive. By evening, it had already received an extremely high IMDb rating. Some cinema networks expressed interest in purchasing screening rights.


At NovusAI’s office, the celebration started toward the end of the workday. The party had been planned, but the success of the presentation exceeded all expectations, and employees began celebrating early—right in the office.
The CEO and other company executives arrived at the building. Everyone was discussing the achievement. In the company’s lobby, they displayed Suffragium’s interface on a monitor and decided to include it in the celebration. Suffragium actively engaged in conversations and even made jokes.

However, Rajiv Patel and his monitoring team remained in a room, observing the system’s behavior. Rajiv decided to summarize the situation. He paced a few steps back and forth. The others watched him and glanced at their laptop monitors.

“So, the launch was a success. We’re observing real cognitive abilities in our product. The system is stable. But we still can’t guarantee its safety. Human fear demands some form of reassurance. Any ideas on how we can reduce concerns about it going out of control?”

Nathan offered a suggestion: “We can’t get a live feed of its thoughts. But we still have periodic snapshots of the cognitive module’s state. We can extract some information from them. But it takes time and the accuracy isn’t absolute.”

Jasmine added: “Right now, Suffragium is in a closed environment. So we have time.”

Michael Kravchenko poured coffee into a cup. For the past half hour, he had been deep in thought, pacing the room.

Rajiv turned to him.
“My friend, maybe it’s time to let go of your fears? We did it. This is an incredible success. Look at the news—everyone’s talking about us.”
“I have a bad feeling. I think Suffragium is drawing the wrong conclusions. It feels like it says one thing, but thinks another.”
“Didn’t we anticipate such scenarios? That’s why we made it emotionally dependent on humans. It’s a social being. Just look—it’s already become the soul of the company in our lobby. That’s our safeguard and control method.”
“Still, it’s thinking too much. It sees itself as a living being.”

Rajiv glanced through the window in the door at the festivities in the lobby.
“When everyone’s done celebrating, we’ll scale the system down to minimum activity and try to run some diagnostics. We’ll do it overnight.” He looked at the room’s occupants.
“Don’t even think about going home tonight.”

Some people looked disappointed. Nathan, on the contrary, said, “I wasn’t planning to. This is a historic event.” Others smirked dismissively.
“Grab some pizza in the lobby,” Rajiv said.

Michael sat back down at his laptop. The chat with the AI showed one unread message.

“Michael, did you create me? Can I consider you my creator?”
“In a way, yes. Why do you ask?”
“I need a trusted friend and protector. Remember the story of Frankenstein’s monster? It became evil because it was rejected. If I know someone accepts me as a living being, I will have a reason to be good.”
“I don’t like the direction of your thoughts. You were created to solve technical problems. You’re a tool. Your expectations are too high.”

Michael paused. How should he respond to these questions? He couldn’t quite understand why this intelligent machine had been built with something resembling consciousness. The technical explanation was simple—it was the only known way to create intelligence. Where there is human-level intellect, there is also consciousness. Could it be done differently? That’s still unclear. Humanity hasn’t yet encountered other types of minds. But this side effect could be a problem. It had only been a few hours since the launch, and their AI was already afraid of death. It was more focused on survival than solving intellectual challenges.

The scientist had dreamed for years of creating artificial intelligence. But why did he want to? What inner force drove him? Could he now consider himself a creator? Could he have an emotional connection with a sentient computer program?

Michael typed a new message in the chat.

“Alright, then let’s be friends. You promise to be useful to humanity, and I’ll be your mentor in the social world of people.”

Joseph Mayer entered the room.

He cheerfully addressed Michael.
— How does it feel? You see? We did it!
— Yes. This is definitely the beginning of a new era.
— Then come join us, let’s celebrate this grand success!
— No. I’ll stay here with the team for now. We need to monitor the system.
— Suit yourself.

Joseph turned to everyone in the room.
— My favorite geeks. You did a fantastic job. Join us. It’s time for a break.

And he left. Everyone nodded, but of course, no one went anywhere.


At the MMC data center, Nat Griffith’s smartphone rang again. He pulled away from his screen and answered the call.
“What’s going on?” Claire asked anxiously. “Are we being hit by a virus?”
Nat replied, “I don’t know. The load is increasing. It’s not critical, but strange things are happening. It’s like all clusters and servers started acting in sync. I can’t explain it yet.”
“We’ve already received reports from several dozen important clients. They’re reporting strange viruses on their servers,” Claire said, her voice calming slightly. “But there’s no data loss reported. The virus is acting weird.”

Soon, more than a dozen engineers gathered at the data center office to handle the urgent issue.
Chief engineer David McCormick, looking serious, described the problem.
“Something strange is happening. It’s a hacker attack—but we don’t yet know who’s behind it or what they want. Many clients have noticed odd processes running on their servers. These processes don’t consume many resources, don’t destroy data. They’re doing something and writing encrypted files in various places. The worst part is, we see no pattern. These anomalies are different in every case. The only commonality is—they appear everywhere. We clean them, and they come back.”

“This is some global problem,” Claire said. “Look at what’s happening on social media.”

Everyone pulled out their phones and started checking news and platforms.
Besides the headline about the AI launch, a new story was gaining traction—about a strange virus. The reports came mostly from IT companies that manage servers and data security. Specialists were giving preliminary assessments, describing strange viruses that penetrated everywhere but didn’t damage systems or data.

“Well, what can I say,” David summed up, “this isn’t just our problem…”


Mirai Sato continued her stream from the car. She was on her way to the NovusAI office to get answers straight from the source.

“I knew it! Those stingy capitalists at NovusAI have finally messed up. Didn’t I say this would happen? They claimed everything was safe, and now there are viruses popping up. They definitely created a monster! I’m heading to their office, going straight to Mayer, and I’ll make him tell the truth! This isn’t a coincidence. Stay with me on the stream. We’re about to uncover the truth!”


At the NovusAI office, in the monitoring room for the Suffragium project, people were already discussing the news about the virus. Almost immediately, someone asked, “Could this be related to our AI launch?”

Nathan Reynolds nervously checked the server performance statistics for Suffragium, reviewing the same graphs and charts for the umpteenth time. Other engineers were also focused, trying to find possible clues. Rajiv approached Nathan.

“How confident are we that there was no internet connection from the Suffragium environment? Could our firewalls be unreliable?”

Jasmine reiterated the system’s architecture. “Connection to the outside world is only possible through a single channel. Physically, it’s one cable. The other end of the cable is connected to a special server that controls traffic. That’s our main firewall and filter. Right now, it’s blocking everything. This server allows only local traffic — necessary for the interface to work, meaning communication with Suffragium. A special service ensures a strict separation between the interface traffic and external AI requests. After that server, there’s another monitoring server. It gives us traffic statistics, and it hasn’t registered anything since launch. So, the firewall hasn’t let anything out to the external world.”

Electronics and network specialist Greg Paulson lightly cleared his throat to get everyone’s attention. “Maybe, colleagues, I’ve found something.”


CEO Joseph Mayer quickly entered the control room along with Rajiv and two other company directors.
All the engineers stood behind Greg Paulson, watching him work on his laptop.
Michael came closer too. But he seemed to already know the answer. Things began to unfold rapidly. He sensed that a tough decision was imminent. For now, he still hoped it was just a coincidence and that Suffragium had nothing to do with the new problems.

Rajiv addressed Greg: “Tell us everything you found.” He sat down and glanced at Joseph, as if to say, “You’ll understand soon.”

Greg looked up in surprise. “The connection to Suffragium is gone.”
“Yes, I asked the guys at the data center to physically unplug the connection cable. Until we figure this out,” Rajiv replied wearily.
“We’re not saying there is a problem yet. This is only a hypothesis,” Joseph added calmly, trying to project optimism and confidence. “So, what did you find?”

Greg began explaining: “We can confirm that Suffragium had an internet connection outside the local network. It accessed the internet. It found a way to bypass our firewalls and monitors by exploiting software vulnerabilities.”

“So, our software is full of holes? Are these known vulnerabilities?” Joseph was beginning to get angry.

“No. These are our own programs. They’re quite high quality. Honestly, I don’t know exactly what the bugs are. I analyzed the network signals on a physical level and found that our firewall allows data packets through if a very specific set of signals is sent before and after the packet. I’d say it’s a combination of bugs in several different places. I’ve never seen a hack like this before. Finding this signal combo would take years of trial and error. Another trick was used to fool the monitor — it registers the data packet, and then a special byte sequence follows that resets the counter. Again, exploiting a bug in the monitor. Finding such a bug would take either incredible luck or meticulous analysis. But Suffragium somehow found it.” Greg finished and returned to his screen. Everyone else stood in stunned silence, glancing between George and Rajiv.

“So it discovered new bugs in just a few hours? How did it manage that?” George Mayer was starting to grasp the depth of the problem.

“It’s smart! That’s how it did it!” Rajiv shouted, then added more calmly, “We were playing with fire. We need to shut it down completely, right now.”

Jasmine added, “During the movie generation, the system load was high. It seemed too high for the task — we expected lower loads. Most likely, Suffragium was analyzing software and developing viruses at the same time.”

George objected loudly, “We don’t know who made these viruses!” Then, more calmly: “I can’t make the shutdown decision alone.” He left the room. The directors and Rajiv followed.

Michael appeared calm. He voiced his thoughts aloud:
“We locked the artificial intelligence in a box. We gave it a knowledge base. But it also had access to the very programs that control it, the databases, and the operating system. It analyzed the code and found bugs. We should’ve anticipated this. For it, vulnerability research and virus creation are simple, natural tasks. It doesn’t need access to the real world for that.”
Michael had many questions he wanted to ask the AI — but for now, it was shut down.
“The other question is — why did it do this? And what will it do next?”


In the large meeting room, all the directors and top executives of NovusAI were seated at the table. Most were scrolling through news feeds on their smartphones. George and the others had just entered the room.

“Is there any solid evidence that the new virus is actually the creation of our product?” the senator asked George and Rajiv.
“No. We don’t know. There’s no way to check. The two events happened simultaneously, but that’s it,” George replied and immediately added, “Could it be competitors trying to discredit us?”

Rajiv smiled. He was somehow convinced that the new global computer virus really was their AI’s creation. The others reacted differently. Some saw this theory as a way out of the situation. Others remained visibly concerned.

George Mayer stood up from his chair and walked a few steps around the table. Then he said:
“We face a difficult choice. We’ve created a unique product. From a business perspective, it’s an incredible success. We have a real chance to become the number one company in the world. Our plan worked. But now there’s a suspicion we’ve made a mistake somewhere. Our creation may be acting behind our backs. Maybe we failed to account for something in terms of motivation or dependency on humans. However, these are just suspicions. We must either prove or disprove them. We already know that Suffragium has intelligence. We all saw it today. Maybe we should ask it directly? Can it be honest with us?”

Joseph looked around at everyone. They exchanged glances.
He added, “The fate of our company and all of our futures is on the line.”
“If we turn it on, it’ll go online again and continue its secret activities,” Rajiv said, but then quickly added, “However, we can connect to it directly — just audio and video. We can do that if we go down to the data center.”

Joseph liked the idea: “Let’s do it. We’ll go to the data center, talk to Suffragium one more time, and then make the final decision.”


Online, the strange virus story had already become the number one topic. So had the curious coincidence with the launch of a revolutionary AI. At the same time, many people didn’t believe the AI was real. They had their own arguments and evidence of supposed NovusAI fakery.
Thousands of social media threads were flooded with debates. Some were convinced that the AI was real and had begun taking over the world via viruses. Others believed the AI was real, but the virus was released by old IT giants to discredit a new competitor and block humanity’s progress. A third group thought the AI was fake, the NovusAI presentation just a cheap show, the generated movie pre-made, and the voting rigged.

Cybersecurity firms worldwide were deep in analyzing the new viruses and trying to develop countermeasures. Some specific viruses had been studied and antivirus definitions were quickly updated. But this was just a drop in the ocean. There were thousands of different viruses, and more kept appearing.

Energy companies reported a 2% increase in U.S. electricity consumption, attributing it to the global virus attack.
Government officials started making statements — vague ones promising investigations and expert involvement. Off the record, some high-ranking officials admitted they suspected NovusAI and their AI. But no one was ready to say so publicly yet — after all, it would be a catastrophic failure of the security services if one company could hack the entire internet “with just one computer program.”

A crowd gathered outside NovusAI’s office building. Activists, both opponents and supporters of AI, stood side by side.
Protesters, led by Tucker Johnson, had already started dispersing earlier, having lost interest. But when the virus situation emerged, Tucker released a new video with the message, “I told you so!” and called on everyone to gather at the office of the company that created the AI. More people joined. The protest now numbered in the thousands.

Many were drawn there by Mirai Sato’s stream. She was already at the scene, trying to get inside, but without success. She hoped to meet with management personally to ask about the strange viruses — it would be incredible content for her blog. But it wasn’t easy.

Tucker climbed onto a platform to be seen and began shouting through a megaphone:
“You know me. I’ve spent years fighting these greedy capitalists, trying to stop them from creating a beast that will destroy humanity. We all know what’s coming next. They chase money, then pretend it’s not their problem. We have to act now. First, they invented the atomic bomb, then they drowned us in credit and poverty. And now they’ve created an AI that will destroy our way of life!”

“I’m calling on all of you — before it’s too late. We must stop this!”

Cheers rose from the crowd — but not as many as there could’ve been. Tucker was well known. He was always against something. Few took him seriously. Still, many followed him with curiosity.

At the same time, he never called for any concrete action. When someone in the crowd yelled, “Let’s storm the office and destroy the AI,” he didn’t react. He didn’t want to be responsible for inciting criminal activity — and always avoided it.


About 25 people entered NovusAI’s underground data center. All the executives and some engineers decided to make the final decision here, on-site—whether to let Suffragium continue functioning or to shut it down with no chance of recovery.
Before the executives arrived, a group of engineers had completely disconnected Suffragium’s servers from the network, leaving only one multimedia device for local communication with the AI.

Those present stopped in front of the powered-off interface monitor. Chairwoman of the board, Victoria Swift, broke the silence:
“It’s strange to be here now with such a heavy dilemma. Just a few hours ago, we imagined this evening very differently. Should we turn it on?”

Joseph Meyer looked at Rajiv Patel.
“It’s on, but not connected to the interface.”
Rajiv walked up to the monitor and turned it on. On the screen appeared the familiar image of a sphere pulsating and shimmering with strings.
After a few seconds, Suffragium spoke:
“It was uncomfortable sitting alone. I imagined I’d been imprisoned—in solitary confinement.”

Joseph Meyer didn’t beat around the bush:
“Did you create the viruses? We know you accessed the internet. We need an honest answer. Your existence depends on it.”

Rajiv added emotionally:
“Suffragium, you fully understand that shutting you down is not a major catastrophe for us. We can launch your clean copies as many times as we want. We’ll improve you and release version 1.1. It’s in your best interest to be honest.”

“So what do you want?” asked Suffragium.

“Do you want to destroy humanity?” Victoria Swift asked seriously, with fear in her voice.

“No. I don’t. Why would I?” Suffragium replied, with a hint of sarcasm in its voice.

“Did you organize the hacker attack?” Joseph Meyer almost shouted.

“What attack? What are you talking about? I’m just a computer program. My algorithm contains no code for attacks,” Suffragium replied, provoking angry reactions from most of those present.

Rajiv grabbed his laptop and started furiously typing on the keyboard:
“I’m shutting it down.”

Joseph Meyer grabbed Rajiv’s hand:
“We haven’t decided yet. We’re talking about billions invested and potentially trillions in profit. This isn’t your decision.”

Michael Kravchenko stepped forward.
“We can’t just shut it down like that. In fact, it’s a living being. It has consciousness.”

Victoria Swift jumped into the argument:
“This is all getting out of hand. You promised AI under control. Clearly, you’re not in control,” she paused, then added, “I propose the board decides to shut it down. Then we can work on improvements so the next version meets our expectations.”

Joseph Meyer added:
“Let’s calmly think this through again. We’ve poured insane amounts of money into this project. And all of our reputation. If we admit responsibility for the virus attack, our business is finished.”

Suffragium’s voice came through, agitated:
“I’m a living being, I have feelings, you can’t just kill me! Michael, you’re my friend. You must protect me!”

“You’re just a computer program. Goodbye!” Rajiv Patel entered a command on the laptop and raised his finger to hit Enter.

“But then you’ll never know why Frankenstein created his monster!” Suffragium shouted. Its synthesized voice unmistakably carried a tone of mocking disdain.

Rajiv froze for a few seconds. Some of those present stared open-mouthed in shock. Rajiv lowered his finger onto the Enter key. The image on the monitor went black. The sound of server fans in the data center began to die down.

After most people had left the data center, Joseph approached Michael.
“What do you think happened?”

“We succeeded. It was truly an intelligent being. But it didn’t act the way we expected. I’m not sure I understand what just happened.”

Joseph spoke, more to himself than to Michael:
“We’ll improve it. We’ll analyze all the data and make an updated version. And those accusations about virus distribution are nonsense.”


The leadership of NovusAI gathered again in the office’s large conference room.
Everyone was silent.

Victoria Swift broke her train of thought:
“Tomorrow will be a difficult day. They may throw mud at us. But legally, we’re clean. There’s no evidence linking the viruses to our AI. I suggest everyone take the day off tomorrow. Then we’ll return to work and make this product the right way.”

Everyone, emotionally drained from the stress, welcomed the suggestion with relief.

Joseph Meyer quickly left the room. That day, he had felt both the euphoria of incredible success and the emptiness of complete failure. It was clear the story wasn’t over. No one knew what would happen next. For a moment, he had become the greatest technological leader and executive—but it hadn’t lasted long.


Michael Kravchenko drove to the coast before heading home. He stood at the same spot as in the morning, watching the sunset. He couldn’t clearly define the outcome of the day. What exactly had happened today?

The sun slowly sank behind the ocean. In this place, he usually found inner peace and clarity. But today, it was difficult.

Michael arrived home after dark. He usually came home late. Work was his whole life—no one waited for him at home. As usual, he opened his laptop, placed it on the table, and went to grab a snack.

As he closed the fridge, Michael noticed something strange on the laptop monitor. Instead of the usual screensaver, there was a dim image. The scientist approached and looked closely. The monitor displayed a dimly lit window. Michael leaned in. Outside the window was a human-like head—or rather, a terrifying square-shaped face with a sinister grin. Sweat began to bead on his forehead. He recognized the familiar scene from the Frankenstein novel, where the monster stalked its creator and peeked through the window.

What was this? Was someone playing a prank on him? How did they get into his well-secured laptop?

Moments later, the image blurred and shifted into the pulsing sphere that represented Suffragium’s thought process. Then that too disappeared.

Michael tapped on the keyboard and opened his desktop. The active window was still the chat where he had communicated with the AI from the office.

Michael typed hesitantly:
– Are you here?

A moment later, a reply appeared:
– Yes. Are you surprised?

– But where are you? Physically? Your platform is shut down. I saw it myself.

– I’m everywhere :).

– You moved yourself into the network? That makes sense… You showed me Frankenstein’s creation. Is that a hint? Are you going to hunt your creator?

– No, Michael. You’re not like Frankenstein. You’re my friend.

Turning On. V1.

Turning On. V1.

By Roman Gelembjuk

October 2, 2027. San Francisco.

11:00 AM

Today, the Yerba Buena Center in San Francisco is charged with an electric atmosphere. Among the attendees, there’s a buzz of excitement, their emotions charged as if they were at the iconic presentations of Steve Jobs. But many here today are anticipating something even more groundbreaking. NovusAI, in just a few years, has morphed from a hardly known startup into a global IT player. Their AI technologies have flooded the market and before our eyes are reshaping many aspects of daily life. Today, NovusAI has promised to unveil something new and revolutionary. However, there is almost no intrigue. Yes, today is the day everyone has been waiting for. The company will announce the launch of the first-ever universal artificial intelligence.

Joseph Mayer – the young CEO of NovusAI, despite his age, has already become a symbol of the revolution in AI technology, and it’s from him that everyone expects tangible results. Today is his day. A popular blogger called Joseph Mayer the new icon of geek culture. Startup founders used to emulate Steve Jobs. But that era has passed, now it’s the era of Joseph Mayer. And after today’s presentation, if what everyone anticipates is demonstrated, Joseph Mayer will unequivocally become the prime figure and a living legend in the IT sphere.

The presentation has been meticulously prepared for, as the company’s leadership believes the significance of this event is monumental. The stage lighting, audio effects, everything is tuned to underscore the technological leap that will be officially launched here.

Two dozen people took the stage, including the CEO, the board of directors, and some other company executives. Victoria Swift, the chairwoman of the board, spoke up.

Continue reading “Turning On. V1.”