We all know that artificial intelligence (AI) is becoming more popular in greater areas of our work. As the Director of Professional Services at vcita, I have witnessed firsthand the profound impact of AI on various industries, revolutionizing the way businesses operate and increasing operational efficiency to unprecedented levels.
AI is reshaping traditional paradigms and unlocking new possibilities for organizations across the globe. From streamlining mundane tasks to enabling data-driven decision-making, AI has emerged as a transformative force. AI tools are empowering businesses to optimize processes and enhance customer experiences. In this article will share my thoughts on the world of AI, how I use it to make processes more efficient, risks, benefits, and even some predictions for the future.
How does AI improve operational efficiency?
According to McKinsey, a whopping 62% of occupations can automate at least 30% of their tasks. And Forbes’, How Businesses Are Using Artificial Intelligence in 2023, uncovered what business owners feel about AI:
- Saves money (59%)
- Decreases response times (53%)
- Prevents mistakes (48%)
- Improves decision-making (44%)
- Streamlines job processes (42%)
So how have I been using AI in my day to day?
AI in my day to day
Meetings
First up, meetings. For many of us, it’s the bane of our existence. They often eat up precious time, keeping us away from the real work. I’ve found a GPT-powered meeting recorder called Speak that transcribes and summarizes meetings for me. No more racking my brain trying to remember everything discussed, or my action items a day later. tl;dv is a game-changer. It not only summarizes the meeting but it also highlights the key points and takes notes for me, allowing me to fully engage and focus on the conversation during the meeting.
Presentations
Presentations are another time-consuming task. Slides.ai or Canva are giant time-savers. I just copy and paste my text, and let it do its thing. It generates fantastic presentations for me. No more slide-by-slide struggle, which as many know, can take hours.
Development
Now, let’s talk development. ChatGPT has become my go-to for automating code generation and bug testing. It’s mind-blowing how seconds with ChatGPT can save me and my team hours of work. Whether it’s creating data models or boilerplate code, ChatGPT has my back. (See my example below.)
Research
I learn from experimenting. I also learn from reading. That’s why AI resources are great to help me learn something new in minutes rather than months. When I come across tl;dr articles, even complicated ones, I toss them into the AI generator and ask it to summarize for me. If the article is particularly complex, I’ll ask ChatGPT to break it down like I’m a 10-year-old. I save precious reading time and I grasp the main point, no matter how intricate the article is.
AI in development
Let me share a cool example where ChatGPT saved the day, no, the week, actually.
Not long ago, my team received a request to complete an SSO integration with a different mechanism than we’d never encountered. Normally something like this would take a few days of research and time to play around with the mechanism to understand it better. I quickly turned to ChatGPT, told it about myself and what kind of expertise it should emulate, and asked it to provide me a code sample in NodeJS.
Here is the prompt I wrote for ChatGPT:
My team needs to develop an SSO integration with a new system that uses SMAL SSO mechanism, can you provide a code example of how this function should look in NodeJS?
Here is ChatGPT’s answer:
“Here’s a step-by-step guide and a code example to set up SSO using SAML with Node.js:
- Install Required Packages
npm install express session passport passport-saml
- Set Up Express Server & Middleware
const express = require(‘express’);
const session = require(‘session’);
const passport = require(‘passport’);
const SamlStrategy = require(‘passport-saml’).Strategy;
const app = express();
app.use(session({
secret: ‘some-secret’,
resave: false,
saveUninitialized: true
}));
app.use(passport.initialize());
app.use(passport.session());
- Configure SAML Strategy
passport.use(new SamlStrategy(
{
entryPoint: ‘https://idp.example.com/login’, // URL for your Identity Provider’s SSO service
issuer: ‘https://sp.example.com/metadata’, // URL for your Service Provider metadata
callbackUrl: ‘https://sp.example.com/login/callback’, // Where the IdP should send the SAML assertions
privateCert: fs.readFileSync(‘./path_to_your_private_key.pem’, ‘utf-8’), // If you have encryption in use
decryptionPvk: fs.readFileSync(‘./path_to_your_private_key.pem’, ‘utf-8’) // If you have encryption in use
},
function(profile, done) {
const user = {
id: profile.nameID,
nameIDFormat: profile.nameIDFormat,
// Add more attributes as needed
};
return done(null, user);
}
));
passport.serializeUser(function(user, done) {
done(null, user);
});
passport.deserializeUser(function(user, done) {
done(null, user);
});
- Add Routes
// Initiates a SAML authentication request
app.get(‘/login’,
passport.authenticate(‘saml’, { failureRedirect: ‘/’, failureFlash: true }),
function(req, res) {
res.redirect(‘/’);
}
);
// Endpoint to receive the SAML Assertion
app.post(‘/login/callback’,
passport.authenticate(‘saml’, { failureRedirect: ‘/’, failureFlash: true }),
function(req, res) {
res.redirect(‘/’);
}
);
app.get(‘/logout’, (req, res) => {
req.logout();
// Add your logout logic here, like redirecting to IdP logout endpoint
res.redirect(‘/’);
});
- Run the Server
app.listen(3000, () => {
console.log(‘Server is running on http://localhost:3000’);
});
“
ChatGPT did about 40% of the work, leaving implementation of the integration as the only thing my team needed to do, and they did it successfully I might add. We completed the project in 2 days instead of a full week of work.
💡Here’s what I recommend for all developers and engineers out there: use ChatGPT for the research, and then, for more complex requests, collect the information and see what the next steps are. Finally, use GPT Engineer and provide it with the relevant prompt. Then, consider letting it handle 50% or more of the actual development.
Staying updated on AI
To stay updated on AI innovations, I recommend reading research articles and using AI tools like ChatGPT to summarize the content for you. Just like I mentioned above. It’s a brilliant way to stay efficient and knowledgeable in minutes.
Reading really is the name of the game.
If research articles
Here are a few people I follow on social media who explain new trends and ideas in the AI landscape thoroughly:
- Marc Köhlbrugge focuses on the latest trends and recommends various tools monthly. He also loves to share other people’s posts that add value to his audience.
- Barsee also talks about the latest trends and information on AI, only from a different angle and POV.
- Paul Couvert is the Founder of Answera, an AI tool for support teams, who started out educating the masses about the power of AI.
- The AI Solopreneur is becoming more and more popular lately and has a newsletter that has updates on the latest news about AI.
Of course, do not forget to follow openAI, Meta AI, Bing, and Google AI because, at the end of the day, they are the trendsetters and pioneers of this AI trend.
Now you have your homework.
The caveat: data privacy
Let’s talk about data privacy—the elephant in the room. As AI relies on data, protecting privacy becomes crucial, especially when dealing with sensitive information such as personal, financial, tech or health-related data. Currently, anything you input into an AI platform will be saved and usable by the public. For example, if I were to input vcita code into Chat GPT (that’s not public knowledge), the code would then be available for all to access.
AI systems are designed to process and analyze vast amounts of data, so organizations must establish robust data governance frameworks and adhere to strict privacy regulations. The challenge lies in developing AI models that can efficiently utilize data while preserving user privacy.
Here are some of the benefits and risks:
Benefits of AI with data privacy
- Anonymization: AI algorithms can be used to remove personally identifiable information from datasets. This can create a great knowledge hub accessible by the masses. Take healthcare professionals. They can share symptoms of their patients, without sharing the patient’s personal information, and gain access to the know-how of millions of other doctors who did the same and described their treatments.
- Data classification: AI can automate sensitive data tagging, enabling tailored privacy safeguards. By tagging the data, I can train the AI algorithm to filter out sensitive data automatically so that users can access the data more efficiently.
- Access control: AI can enforce precise data access rights, leveraging behavioral analysis for enhanced protection. Any company can use this type of protection. It helps automatically give permissions and access to certain employees based on their needs, roles, etc. Imagine an employee leaves your organization. You’d need access to all their knowledge and projects they were working on in the company. AI can keep track of this and automatically assign the data to the employee’s replacement.
It is important to note that while AI can contribute to data privacy, it also raises concerns about potential privacy risks.
Risks of AI with data privacy
- Inferred data: AI’s power to deduce sensitive information from seemingly innocuous data can jeopardize privacy. For example, AI could use the information tracked from your smartwatch or fitness app to make assumptions about your health condition and then share that information with organizations that will use it to target you with specific ads.
- Bias and discrimination: Unchecked AI can amplify biases, perpetuating discriminatory outcomes. An example here is someone whose financial standing is good, and is looking to take out a mortgage, but comes from a lower socioeconomic area (just for this example), might get rejected because the AI tool gathered so much data about their surrounding circumstances and didn’t take into account their current standing.
- Consent and control: Insufficient consent and user control undermine privacy principles. Bringing this into laments terms: someone signs up to a music streaming service and agrees to the terms and conditions without really reading them (as most of us do). What they didn’t read thoroughly enough was that the Ts&Cs allow for the streaming service to use AI to analyze their listening habits and recommend certain ads based on their mood. This means the information about the listening habits was shared with third party advertisers.
- Re-identification: AI’s ability to combine datasets risks unintentional re-identification and privacy compromise. Say someone participated in a medical study and their health data was anonymized and then published for research purposes. However, an AI system cross-references this dataset with publicly available data from this person’s social media, fitness apps, and other online footprints, so much so that it can now re-identify this person and link them to their previously anonymized data.
To mitigate risks, we must prioritize fairness and embed privacy by design into our systems. We can ensure that AI technologies do not compromise privacy by implementing appropriate safeguards to protect sensitive data throughout the AI lifecycle. For instance, we could limit data collection and retention to only the minimum amount necessary for the AI system’s intended purpose. Additionally, pseudonymizing the data whenever possible can help prevent the identification of individuals.
On the human level, it’s crucial to ensure that team members refrain from sharing sensitive information on open AI platforms that could compromise privacy. By being vigilant about data privacy, we can look forward to a future where AI continues to bring value to the workplace.
The future of AI in the workplace
While I can’t, of course, predict the exact details of how the future will look, one thing is clear: communication will undergo a radical transformation. Thanks to AI.
AI in the workplace promises intriguing transformations in communication. Advancements in natural language processing and voice recognition hold the potential to revolutionize how we interact and collaborate. While the specifics of these advancements are still unfolding, we can anticipate improvements like enhanced connectivity, real-time translation capabilities, immersive virtual environments, and sophisticated personal assistants—all contributing to greater efficiency and productivity.
Here are my predictions:
- Ubiquitous connectivity: AI will leverage 5G/6G and satellite technology for universal connectivity.
- Enhanced real-time translation: AI will enable flawless, real-time translations—and I’m super-excited about this possibility because it will essentially eliminate language barriers.
- Immersive virtual environments: AI, Virtual Reality (VR), and Augmented Reality (AR) will create realistic virtual environments for communication, revolutionizing work, learning, and social interaction.
- Smarter personal assistants: Advanced AI assistants will understand context and preferences to manage communication and enhance productivity.
- Privacy and security: AI will learn to leverage the benefits I listed above to ensure digital communication’s privacy and security, detect threats, manage data privacy, and maintain secure channels.
With advancements in natural language processing and voice recognition, we can expect AI-powered communication tools that will revolutionize how we interact and collaborate. But who can know how this will look? It may be with an evolved smartphone, revamped Google Glass or no device at all! We can only imagine.
One thing’s for sure, it’s in our best interest to start using AI every chance we have.
And to all those that are thinking AI will replace them in their jobs I say this: AI is not a thinking and breathing being, it solely executes commands based on data processing. Learning to use this powerful tool will help you become more efficient in your job, which will definitely open up new doors and opportunities. Bottom line: you need to bring the ideas and let it work for you, not instead of you.
Learn to communicate with your new team member: AI
So, how can you embark on your AI journey? Begin by identifying an area of opportunity or a specific challenge in your workplace—whether it’s summarizing meetings, crafting presentations, or enhancing communication. Once you’ve pinpointed the issue, explore AI tools designed to address it.
Mastering the functionality of these tools may require some trial and error, but the investment in time and effort is well worth the payoff. Consider treating AI like an intern: if you provide clear and precise instructions, the likelihood of receiving the output you desire increases significantly.
As you grow more comfortable with the ideal AI tool for your needs and learn how to effectively interact with it, you’ll be astonished at how much you can achieve. You might even discover that your productivity has doubled!
