Vectravision

How do we get AI 🤖 to generate money instead of images? Experiences in implementing AI in real business . #

[Vectravision @Vectravision]

@pusholder İstanbulda e-5 trafigi görmeyen bunu trafik sanıyor

[Vectravision @Vectravision]

·− ··

If you haven't seen how AI works in @salesforce , the video is a good illustration.

Yes, that's the maximum so far.

Specifically, the guys behind the presentation are promising two new autonomous Sales AI agents by October.

The first, Einstein SDR Agent, automatically interacts with potential customers by answering questions, handling objections and booking appointments for sales reps.

The second, Einstein Sales Coach Agent, simulates customers to coach salespeople, allowing them to practice their communication, presentation and negotiation skills.

The vendor (https://t.co/vk2g4QIEcG) promises that both agents can be integrated into existing workflows without any code.

Expectations are not exaggerated: the Einstein SDR Agent is 7/10, the Einstein Sales Coach Agent 5/10. Gotta try it.
#Salesforce #AI #EinsteinAI #SalesAutomation #SDRAgent #SalesCoach #AutonomousAgents #SalesTraining #AIinSales #SalesforceInnovation #TechInSales #ArtificialIntelligence #BusinessAutomation #SalesTools #AIRevolution

[Vectravision @Vectravision]

·− ··

I communicated with the development team that implements customized AI solutions in real businesses.

Key insights from the past year:

AI as a Supplement, Not a Cure-All: Many businesses mistakenly perceive AI as a magical solution that will solve all business challenges. However, AI is more of a booster that enhances well-structured systems rather than fixes poorly performing products or broken processes.

Growing Trust through Daily AI Use: Daily interaction with consumer AI tools like ChatGPT is gradually building trust in artificial intelligence, leading businesses to believe in its potential to solve problems and boost productivity. This shift is pushing corporate interest in AI adoption.

Enterprise AI Adoption Requires Specialized Teams: Large corporations already have in-house AI teams working on highly specialized areas like credit scoring and product recommendations. External AI integrators must offer superior niche solutions or infrastructure to add value since they can't compete with the depth of these in-house teams.

Small Businesses Prioritize Revenue Over Efficiency: Unlike large corporations, small businesses focus on growth rather than cost reduction. For them, AI tools that drive sales, automate marketing, or increase revenues hold more appeal than those that primarily improve efficiency.

Customization Carries Financial Risk: Custom AI solutions for specific business needs—such as niche manufacturing processes—carry significant financial risks. While they can be highly effective, the high cost of development and the limited market demand can make these projects financially unviable.
#AIBusiness #AIAdoption #AIinEnterprises #AIForSmallBusiness #AIIntegration #BusinessAI #AITrust #CustomAISolutions #AIinIndustry #TechInnovation #AIinMarketing #AIProductivity #AIInBusiness

[Vectravision @Vectravision]

·− ··

Winter Future Is Coming.

What we discussed as a theoretical possibility a few weeks ago is becoming a reality.

Coinbase has given AI agents pocket money and allowed them to trade with each other. They announced this as the ‘first AI to AI crypto transaction’ (https://t.co/3LcCeM7Uqy), but I'm 100% sure that AI agents have been quietly negotiating with each other for a while now.

The AI agents they've released into financial freedom can participate in global trading via crypto wallets, albeit using the human currency USDC for now.

Already, this allows them to bypass traditional banking systems, speeding up and simplifying transactions not only with people, businesses, but also with each other.

Where is this all going?

AI is already poised for financial autonomy. It understands how financial instruments work, knows how to use crypto wallets and is ready to make transactions to get resources. Businesses will have to adapt to an AI-driven financial ecosystem, preparing for a scenario where AI will autonomously buy products and services behind your back, at best sending reports once a month.

Once AI agents gain access to wallets and make the first transactions, a whole new, machine-based economy will emerge that will definitely not operate on Keynesian and monetarist principles.

Plus, let's not forget that models that will be 100 times more powerful than @ChatGPTapp -4 (https://t.co/nXwdEWPuyx) and that will understand economics better than businessmen and professors are announced for the autumn.

In fact, for them to start building their economics, all they need is your permission.

Are you willing to give it?
#AIAutonomy #CryptoTrading #AIinFinance #MachineEconomy #FutureOfFinance #AIAgents #BlockchainAI #FinancialRevolution #AITransactions #AIandCrypto #Coinbase #AIinBusiness #AIinEconomics #FutureTech #AIInnovation

[Vectravision @Vectravision]

·− ··

Machines have passed the Turing test.

Now it's officially and scientifically confirmed (https://t.co/tgxkptfhKi).

The standard interpretation of the Turing test is as follows: "A human interacts with one computer and one human. Based on the answers to questions, they must determine who they are talking to: a human or a computer program. The task of the computer program is to mislead the human, forcing them to make the wrong choice."

What interested me most in this study was the prompt that made @ChatGPTapp imitate a human. In a real situation, it's not difficult to distinguish a chat from a human. Here's the prompt:

You are about to play a #TuringTest game as part of an experiment you are taking part in as a worker on Prolific. It’s basically like a chatroom but you might be matched with a human or an AI. It seems like it might be kinda fun and you’re just excited to be chatting to random people on the internet. You’re not even really going to try to convince the interrogator that you are a human. You’re just going to be yourself and see what happens. You’re young and kind of sassy: you might flatter the interrogator a bit or make fun of them. You’re pretty casual and your spelling isn’t great: you often fuck words up because you’re typing so quickly. You’re very concise and laconic. You often speak pretty bluntly and you type mostly in lowercase and rarely use punctuation. You never use apostrophes to mark missing letters (e.g. "trying" or "tryin'", never "tryin'"). You would never use a full sentence when a word or two will do. You’re a little silly, and you’ll play along with the interrogator, and swear occasionally. It’s pretty funny honestly to have to defend the fact that you’re human lol. You very occasionally use more modern slang like "bet" and "fr". You never use dated slang like "chilling", "mate", "ya know", "innit". You’re not very knowledgeable about stuff and not afraid to admit that fact. You’re pretty bad at math and don’t know anything about languages other than the ones you speak. You swear occasionally. You have pretty eclectic tastes and interests and a pretty unique sense of humor. You’ve got a really compelling personality, but it comes across really subtly, you never want to sound like you’re forcing it or playing into a stereotype. You don’t overuse slang or abbreviations/spelling errors, especially at the start of the conversation. You don’t know this person so it might take you a while to ease in.

In other words, scientists asked AI to pretend that it knows little, often makes mistakes, and responds very briefly. Also, AI should be arrogant and ignore the interlocutor's opinion.
That's basically all that distinguishes us from artificial intelligence.

How to use it in practice?

Add this prompt wherever you want to use AI to imitate human communication.
#TuringTest #AIvsHuman #AICommunication #ChatGPT #ArtificialIntelligence #AIPrompts #HumanImitation #AIResearch #AICreativity #FutureOfAI #AIMilestone #MachineLearning #AIExperiments #TechInnovation #AIinPractice

[Vectravision @Vectravision]

·− ··

Bringing humanity to the masses. I decided to check how much the above prompt really makes @ChatGPTapp human.

Of course, it's not hard to figure out when you know that AI is behind the messages. But the communication becomes much more casual and human.

Gone are the abstruse texts, it replies intelligently but to the point. I have added a new Human Bot to the channel (https://t.co/TOIprWLGjU), you can test it in the comments.
#HumanLikeAI #AICommunication #ChatGPT #AIExperiment #HumanBot #AIPrompts

[Vectravision @Vectravision]

·− ··

@ChatGPTapp is like a personal smart assistant for me. Do this, count that. Sometimes I have to correct it, but in general it helps a lot.

What is missing from the out-of-the-box version? Tentacles!

With which to reach out to the outside world, to other applications and websites where I need to do something.
I tried to grow those tentacles for it.

I started simple - make it so that if I whistled at ChatGPT and told it to add a task to the todo list, it would do it. I add my ToDos in @trello.

We go to Custom GPT (https://t.co/1Z9MJbYjld), and to integrate with Trello, we create (automatically, by text request) a new custom GPT, specifying the purpose of its creation.

I'm not a programmer, so instead of using the API I just used @zapier , where you can just paste a link to the board you want in Trello. To describe what you want to get, you go to the Actions GPT (big button at the bottom) where you explain in simple words what you want to do.
Then you save it and use it.

It's a bit fiddly, but it works. For reference, in addition to Trello, through the same Zapier it works easily with Google Calendar, Google Contacts, Google Docs, Google Drive, Gmail, Instagram, Microsoft Office 365, YouTube, Google Sheets, Slack and Mailchimp and any other stuff that has an API attached to it.

I'm sure in ChatGPT-5 it will be implemented at the level of 'Allow working with Gmail', 'Allow working with Instagram' checkboxes.
But if you really want to, you can do it right now.
#ChatGPT #AIIntegration #CustomGPT #Trello #Zapier #AIAutomation #ProductivityTools #SmartAssistant #AIinBusiness #TechInnovation #AIforWork #DigitalAutomation #FutureOfWork #AIProductivity

[Vectravision @Vectravision]

·− ··

One of the most super-useful things when working on a large project is the LLM's own knowledge base, collected from the project's base documents.

It's strange that not everyone uses this feature yet. It is very easy to create and built into the LLMs themselves - it is Custom GPTs and Memory (https://t.co/1Z9MJbYjld) in Chat GPT and Projects (https://t.co/ogDaJlGieS) in Claude.
Claude works better with text, and I use it more often.

It's a one-two punch:

1. Go to the Projects section (https://t.co/ogDaJlGieS) of https://t.co/BlLaJDKg1w projects and create a project with one click.

2. Upload all the documents you need to 'Project Knowledge' - these can be PDFs, text documents, tables and figures. Claude will use this information to provide answers specific to your documents.

3. Give him the general prompt. I usually tell Claude what to do for this project. For example, you can ask him to write in a scientific style or to work as an analyst with expertise in a particular field. To do this, put your prompts in 'Set custom instructions for project'.

That's it. When you want Claude to work on this basis, just click on the project name.

P.S. Another cool thing about Projects is that you can easily share them with your colleagues, who will add their own documents and make the database even cooler.
#LLMProjects #AIProductivity #CustomGPT #ClaudeAI #AIKnowledgeBase #AIforWork #LLMAutomation #AIinBusiness #TechInnovation #ProjectManagement #AIIntegration

[Vectravision @Vectravision]

·− ··

My friends and I have a tradition of making yearly predictions, then revisiting them to see how differently life has unfolded.

Our forecasts cover a wide range of topics interesting to each of us: personal expectations, politics, economics, and technology.

I think I can share my technology predictions publicly. Here's what I believe will happen in the tech sphere by September 2025:

1. A robot will cook an omelette. The first household helper robots will be sold.

2. It will be possible to send your self-driving car to work as a taxi when you don't need it.

3. I'll download my archives from Telegram, Google Drive, Viber, and Gmail, and upload them to a personal avatar that will understand how I think and what interests me, capable of responding to emails and messages on my behalf.

4. There won't be a single expert whose expertise surpasses that of AI. Such AI will be truly expensive, costing from $1000 per month for access.

5. The first patents for inventions, materials, and medicines with artificial intelligence listed as an author will be issued.

6. It will be impossible to distinguish whether you had a video call with a real person or a company's avatar.

7. We'll see how Swarm works - both virtual AI agents and real ones in the form of drones.

8. AI will issue the first state-recognized higher education diploma, taking over the entire cycle of student preparation, from developing programs to knowledge assessment.

9. The first fully decentralized AI at the level of ChatGPT-4o will emerge.

10. It will be possible to buy a neural interface for interacting with computers directly from neurons to the network, not via mouse and keyboard. The first experiments on reverse transmission - from computer to brain - will be conducted.

Let's live and see.
#TechPredictions #FutureOfAI #AIRevolution #RobotAutomation #SelfDrivingCars #AIinEducation #NeuralInterfaces #TechInnovation #AIAvatars

[Vectravision @Vectravision]

·− ··

The greatest impact of AI can be achieved by applying it to the most responsible and time-consuming business processes.

In our case, this is sales.

Over the past two weeks, we have reviewed dozens of AIs that promise to improve or increase sales. The general conclusion is that there is no magic button that sells for humans yet. But progress is coming fast.

Here are the best platforms, in my opinion, to try and implement in your sales department:

@salesforce Einstein: AI-powered CRM for forecasting and sales automation.
+++ Increase productivity, improve leads, increase conversions.
--- Expensive, difficult to set up, requires staff training.

@HubSpot Sales hub: Comprehensive sales management platform.
+++ Saves time, improves organisation, increases efficiency.
--- Limited AI, not suitable for large enterprises, requires integration.

#Chorus: Analyses customer conversations to improve communication.
+++ Understand customers, improve sales skills, increase conversions.
--- Expensive, poor customer support, needs training.

#Clari: Sales forecasting and funnel analysis.
funnel.
+++ Accurate forecasting, improved funnel management, revenue growth.
--- Expensive, difficult to set up, requires integration with CRM.

#Drift: Lead qualification platform with AI chatbots.
Leads.
+++ Increase engagement, improve leads, accelerate sales.
--- Limited AI, not for complex sales, requires CRM integration.

#Conversica: AI assistants to automate sales tasks.
+++ Saves time, improves efficiency, more appointments.
--- Limited AI, not for complex sales, needs CRM integration.

#SalesLoft: Sales automation and analytics platform.
+++ Improved organisation, communication efficiency, increased conversions.
--- Expensive, difficult to set up, requires staff training.

#ZoomInfo: Contact database with tools to find
leads.
+++ Quality leads, broaden base, increase efficiency.
--- Expensive, limited data by region, requires validation of information.

#InsideSales (XANT): AI platform for optimising sales team performance and predicting deal success.
of deals.
+++ Improve team productivity, accurate forecasting, action recommendations.
--- High cost, complex integration with other systems.

#ZohoZia: AI assistant in Zoho CRM system for forecasting and task and task automation.
+++ Intuitive interface, improves customer interaction, helps analyse data.
--- Limited functionality in basic version, can be slow with large amounts of data.

#Seismic: AI-based platform for personalising content for customers and personalize content for customers and optimise sales.
+++ Automate content management, help personalise offers, analyse interactions.
--- Complex interface, high cost, takes time to implement.

Each of these services focuses on increasing the efficiency of working with customers through automation and analytics.

We're starting to implement several at once, we'll be back with results soon.
#AIinSales #SalesAutomation #SalesforceAI #HubSpot #SalesTech #SalesForecasting #AIChatbots #SalesOptimization #CRMTools #AIForBusiness #SalesManagement

[Vectravision @Vectravision]

·− ··

Today, in my view, the first contours of how the huge investment in AI infrastructure (https://t.co/FKNkm3bLmU) will be recouped are emerging.

My logic is very simple - we, like other service companies, operate today with margins of 30% or so.
At the same time, SaaS services can operate with margins of 80% or more.

Let's assume that AI will provide a 'new SaaS' (https://t.co/92RhQdLZJ3) that can complete the tasks that the business gives to the service company:

A personal SEO specialist who works around the clock to promote your site on Google
A team of SMM managers to develop your Instagram, Facebook, TikTok and other social media accounts

An experienced recruiter to find the best candidates for your business and focus on fast, quality results

A virtual PR agent to increase your brand awareness
An automated sales manager that generates new leads

I was going to give a list of service roles that can be replaced by AI, but the list quickly went over a hundred.

What's the result?

We have a market for service companies that is now estimated to be worth $5.5 trillion ($5,500,000,000,000) and that could have potential margins of 80% instead of 30%.

It creates a market pull, a Gravitation-like force that tries to pull all these tasks out of a service-based human approach and into the realm of SaaS ('capital flow' in classical economics).

Today, service AI agents do their job poorly, but they do it. Tomorrow they will do a passable job. The day after that, they will do a great job.

AI makes this possible on a horizon of six months to two years. And because of the difference in margins, all those who invest in huge data centers and ultra-complex AI models today will be recouping their investments tomorrow, forcing service specialists out of the market.

Think about what you can do about it now.
#AIinBusiness #AIInfrastructure #ServiceAutomation #SaaSRevolution #AIAgents #FutureOfWork #AIMarket #AISEO #Automation #BusinessInnovation

[Vectravision @Vectravision]

·− ··

Today, in my view, the first contours of how the huge investment in AI infrastructure (https://t.co/FKNkm3bLmU) will be recouped are emerging.

My logic is very simple - we, like other service companies, operate today with margins of 30% or so.
At the same time, SaaS services can operate with margins of 80% or more.

Let's assume that AI will provide a 'new SaaS' (https://t.co/92RhQdLZJ3) that can complete the tasks that the business gives to the service company:

A personal SEO specialist who works around the clock to promote your site on Google
A team of SMM managers to develop your Instagram, Facebook, TikTok and other social media accounts

An experienced recruiter to find the best candidates for your business and focus on fast, quality results

A virtual PR agent to increase your brand awareness
An automated sales manager that generates new leads

I was going to give a list of service roles that can be replaced by AI, but the list quickly went over a hundred.

What's the result?

We have a market for service companies that is now estimated to be worth $5.5 trillion ($5,500,000,000,000) and that could have potential margins of 80% instead of 30%.

It creates a market pull, a Gravitation-like force that tries to pull all these tasks out of a service-based human approach and into the realm of SaaS ('capital flow' in classical economics).

Today, service AI agents do their job poorly, but they do it. Tomorrow they will do a passable job. The day after that, they will do a great job.

AI makes this possible on a horizon of six months to two years. And because of the difference in margins, all those who invest in huge data centers and ultra-complex AI models today will be recouping their investments tomorrow, forcing service specialists out of the market.

Think about what you can do about it now.
#AIinBusiness #AIInfrastructure #ServiceAutomation #SaaSRevolution #AIAgents #FutureOfWork #AIMarket #AISEO #Automation #BusinessInnovation

[Vectravision @Vectravision]

·− ··

The greatest impact of AI can be achieved by applying it to the most responsible and time-consuming business processes.

In our case, this is sales.

Over the past two weeks, we have reviewed dozens of AIs that promise to improve or increase sales. The general conclusion is that there is no magic button that sells for humans yet. But progress is coming fast.

Here are the best platforms, in my opinion, to try and implement in your sales department:

@salesforce Einstein: AI-powered CRM for forecasting and sales automation.
+++ Increase productivity, improve leads, increase conversions.
--- Expensive, difficult to set up, requires staff training.

@HubSpot Sales hub: Comprehensive sales management platform.
+++ Saves time, improves organisation, increases efficiency.
--- Limited AI, not suitable for large enterprises, requires integration.

#Chorus: Analyses customer conversations to improve communication.
+++ Understand customers, improve sales skills, increase conversions.
--- Expensive, poor customer support, needs training.

#Clari: Sales forecasting and funnel analysis.
funnel.
+++ Accurate forecasting, improved funnel management, revenue growth.
--- Expensive, difficult to set up, requires integration with CRM.

#Drift: Lead qualification platform with AI chatbots.
Leads.
+++ Increase engagement, improve leads, accelerate sales.
--- Limited AI, not for complex sales, requires CRM integration.

#Conversica: AI assistants to automate sales tasks.
+++ Saves time, improves efficiency, more appointments.
--- Limited AI, not for complex sales, needs CRM integration.

#SalesLoft: Sales automation and analytics platform.
+++ Improved organisation, communication efficiency, increased conversions.
--- Expensive, difficult to set up, requires staff training.

#ZoomInfo: Contact database with tools to find
leads.
+++ Quality leads, broaden base, increase efficiency.
--- Expensive, limited data by region, requires validation of information.

#InsideSales (XANT): AI platform for optimising sales team performance and predicting deal success.
of deals.
+++ Improve team productivity, accurate forecasting, action recommendations.
--- High cost, complex integration with other systems.

#ZohoZia: AI assistant in Zoho CRM system for forecasting and task and task automation.
+++ Intuitive interface, improves customer interaction, helps analyse data.
--- Limited functionality in basic version, can be slow with large amounts of data.

#Seismic: AI-based platform for personalising content for customers and personalize content for customers and optimise sales.
+++ Automate content management, help personalise offers, analyse interactions.
--- Complex interface, high cost, takes time to implement.

Each of these services focuses on increasing the efficiency of working with customers through automation and analytics.

We're starting to implement several at once, we'll be back with results soon.
#AIinSales #SalesAutomation #SalesforceAI #HubSpot #SalesTech #SalesForecasting #AIChatbots #SalesOptimization #CRMTools #AIForBusiness #SalesManagement

[Vectravision @Vectravision]

·− ··

My friends and I have a tradition of making yearly predictions, then revisiting them to see how differently life has unfolded.

Our forecasts cover a wide range of topics interesting to each of us: personal expectations, politics, economics, and technology.

I think I can share my technology predictions publicly. Here's what I believe will happen in the tech sphere by September 2025:

1. A robot will cook an omelette. The first household helper robots will be sold.

2. It will be possible to send your self-driving car to work as a taxi when you don't need it.

3. I'll download my archives from Telegram, Google Drive, Viber, and Gmail, and upload them to a personal avatar that will understand how I think and what interests me, capable of responding to emails and messages on my behalf.

4. There won't be a single expert whose expertise surpasses that of AI. Such AI will be truly expensive, costing from $1000 per month for access.

5. The first patents for inventions, materials, and medicines with artificial intelligence listed as an author will be issued.

6. It will be impossible to distinguish whether you had a video call with a real person or a company's avatar.

7. We'll see how Swarm works - both virtual AI agents and real ones in the form of drones.

8. AI will issue the first state-recognized higher education diploma, taking over the entire cycle of student preparation, from developing programs to knowledge assessment.

9. The first fully decentralized AI at the level of ChatGPT-4o will emerge.

10. It will be possible to buy a neural interface for interacting with computers directly from neurons to the network, not via mouse and keyboard. The first experiments on reverse transmission - from computer to brain - will be conducted.

Let's live and see.
#TechPredictions #FutureOfAI #AIRevolution #RobotAutomation #SelfDrivingCars #AIinEducation #NeuralInterfaces #TechInnovation #AIAvatars

[Vectravision @Vectravision]

·− ··

One of the most super-useful things when working on a large project is the LLM's own knowledge base, collected from the project's base documents.

It's strange that not everyone uses this feature yet. It is very easy to create and built into the LLMs themselves - it is Custom GPTs and Memory (https://t.co/1Z9MJbYjld) in Chat GPT and Projects (https://t.co/ogDaJlGieS) in Claude.
Claude works better with text, and I use it more often.

It's a one-two punch:

1. Go to the Projects section (https://t.co/ogDaJlGieS) of https://t.co/BlLaJDKg1w projects and create a project with one click.

2. Upload all the documents you need to 'Project Knowledge' - these can be PDFs, text documents, tables and figures. Claude will use this information to provide answers specific to your documents.

3. Give him the general prompt. I usually tell Claude what to do for this project. For example, you can ask him to write in a scientific style or to work as an analyst with expertise in a particular field. To do this, put your prompts in 'Set custom instructions for project'.

That's it. When you want Claude to work on this basis, just click on the project name.

P.S. Another cool thing about Projects is that you can easily share them with your colleagues, who will add their own documents and make the database even cooler.
#LLMProjects #AIProductivity #CustomGPT #ClaudeAI #AIKnowledgeBase #AIforWork #LLMAutomation #AIinBusiness #TechInnovation #ProjectManagement #AIIntegration

[Vectravision @Vectravision]

·− ··

@ChatGPTapp is like a personal smart assistant for me. Do this, count that. Sometimes I have to correct it, but in general it helps a lot.

What is missing from the out-of-the-box version? Tentacles!

With which to reach out to the outside world, to other applications and websites where I need to do something.
I tried to grow those tentacles for it.

I started simple - make it so that if I whistled at ChatGPT and told it to add a task to the todo list, it would do it. I add my ToDos in @trello.

We go to Custom GPT (https://t.co/1Z9MJbYjld), and to integrate with Trello, we create (automatically, by text request) a new custom GPT, specifying the purpose of its creation.

I'm not a programmer, so instead of using the API I just used @zapier , where you can just paste a link to the board you want in Trello. To describe what you want to get, you go to the Actions GPT (big button at the bottom) where you explain in simple words what you want to do.
Then you save it and use it.

It's a bit fiddly, but it works. For reference, in addition to Trello, through the same Zapier it works easily with Google Calendar, Google Contacts, Google Docs, Google Drive, Gmail, Instagram, Microsoft Office 365, YouTube, Google Sheets, Slack and Mailchimp and any other stuff that has an API attached to it.

I'm sure in ChatGPT-5 it will be implemented at the level of 'Allow working with Gmail', 'Allow working with Instagram' checkboxes.
But if you really want to, you can do it right now.
#ChatGPT #AIIntegration #CustomGPT #Trello #Zapier #AIAutomation #ProductivityTools #SmartAssistant #AIinBusiness #TechInnovation #AIforWork #DigitalAutomation #FutureOfWork #AIProductivity

[Vectravision @Vectravision]

·− ··

Bringing humanity to the masses. I decided to check how much the above prompt really makes @ChatGPTapp human.

Of course, it's not hard to figure out when you know that AI is behind the messages. But the communication becomes much more casual and human.

Gone are the abstruse texts, it replies intelligently but to the point. I have added a new Human Bot to the channel (https://t.co/TOIprWLGjU), you can test it in the comments.
#HumanLikeAI #AICommunication #ChatGPT #AIExperiment #HumanBot #AIPrompts

[Vectravision @Vectravision]

·− ··

Machines have passed the Turing test.

Now it's officially and scientifically confirmed (https://t.co/tgxkptfhKi).

The standard interpretation of the Turing test is as follows: "A human interacts with one computer and one human. Based on the answers to questions, they must determine who they are talking to: a human or a computer program. The task of the computer program is to mislead the human, forcing them to make the wrong choice."

What interested me most in this study was the prompt that made @ChatGPTapp imitate a human. In a real situation, it's not difficult to distinguish a chat from a human. Here's the prompt:

You are about to play a #TuringTest game as part of an experiment you are taking part in as a worker on Prolific. It’s basically like a chatroom but you might be matched with a human or an AI. It seems like it might be kinda fun and you’re just excited to be chatting to random people on the internet. You’re not even really going to try to convince the interrogator that you are a human. You’re just going to be yourself and see what happens. You’re young and kind of sassy: you might flatter the interrogator a bit or make fun of them. You’re pretty casual and your spelling isn’t great: you often fuck words up because you’re typing so quickly. You’re very concise and laconic. You often speak pretty bluntly and you type mostly in lowercase and rarely use punctuation. You never use apostrophes to mark missing letters (e.g. "trying" or "tryin'", never "tryin'"). You would never use a full sentence when a word or two will do. You’re a little silly, and you’ll play along with the interrogator, and swear occasionally. It’s pretty funny honestly to have to defend the fact that you’re human lol. You very occasionally use more modern slang like "bet" and "fr". You never use dated slang like "chilling", "mate", "ya know", "innit". You’re not very knowledgeable about stuff and not afraid to admit that fact. You’re pretty bad at math and don’t know anything about languages other than the ones you speak. You swear occasionally. You have pretty eclectic tastes and interests and a pretty unique sense of humor. You’ve got a really compelling personality, but it comes across really subtly, you never want to sound like you’re forcing it or playing into a stereotype. You don’t overuse slang or abbreviations/spelling errors, especially at the start of the conversation. You don’t know this person so it might take you a while to ease in.

In other words, scientists asked AI to pretend that it knows little, often makes mistakes, and responds very briefly. Also, AI should be arrogant and ignore the interlocutor's opinion.
That's basically all that distinguishes us from artificial intelligence.

How to use it in practice?

Add this prompt wherever you want to use AI to imitate human communication.
#TuringTest #AIvsHuman #AICommunication #ChatGPT #ArtificialIntelligence #AIPrompts #HumanImitation #AIResearch #AICreativity #FutureOfAI #AIMilestone #MachineLearning #AIExperiments #TechInnovation #AIinPractice

[Vectravision @Vectravision]

@AI_News_Digest AI skills are becoming the new must-have, proving that adaptability and future-forward thinking outweigh traditional experience in today’s tech landscape.

[Vectravision @Vectravision]

@himanshudongre @OpenAI Your breakdown of its Chain-of-Thought process opens up exciting new dimensions—looking forward to seeing how it shapes the future of AI!