TL;DR – What You Need to Know in 30 Seconds

• Pinpoint your core pain points so you know exactly what you want your agent to solve.
• Choose the right tool: collaborative, automation, or social AI agents have distinct roles and strengths.
• Tread carefully: build your agent step-by-step and pilot it before a full-scale launch.
• Keep it human: no matter how advanced your agent is, you’ll need oversight and guidance to make it thrive.

Can AI Agents Can Really Transform Your Business?

Could 2025 Be the Year of AI Agents?

These brilliant, autonomous systems are set to take centre stage by making decisions and executing tasks without the need for constant human prodding. It’s no wonder that savvy business leaders across Asia (and beyond!) are champing at the bit to integrate them into their workflows. After all, who wouldn’t want to delegate time-consuming tasks to an unfailingly efficient digital helper?

But, as with any major tech leap, bringing AI agents into your business isn’t something you do at the drop of a hat. Trust me—my team and I recently built and launched our very own AI agent. The learning curve was steep, and along the way, I discovered a series of crucial steps you can’t afford to skip if you want to do it right. So, get comfy, and let’s explore how these pint-sized powerhouses can transform your operations.

Identify Your Business Needs

Let’s be honest: we’ve all been witnessing peak AI mania this past year. From wacky (and sometimes pointless) inventions like the $3,500 AI-enabled toaster to questionable ChatGPT-authored blog posts, it’s been a rollercoaster. Yet for all the hype, AI genuinely has reimagined the way we work. Its ability to process massive volumes of data and automate complex processes is beyond impressive. In fact, it can be downright revolutionary.

Still, in the rush to try every fancy new AI tool on the market, it’s easy to lose sight of what you actually need.

Phu Nguyen’s point perfectly illustrates the potential scope of AI empowerment in the workplace. But remember, before you throw an agent at every minor problem, sit down and identify the core challenges you face.

Whether you’re looking to speed up customer service response times, reduce operational bottlenecks, or optimise your supply chain, a clear understanding of the problem you need to solve is vital. Otherwise, you’ll risk investing in an agent that creates more headaches than it cures.

Pick Your AI Agent

Here’s the thing: not all AI agents are created equal. Much like your toolbox at home, you’ve got different gadgets for different jobs. You wouldn’t reach for a hammer if you needed to tighten a screw, would you? So, once you’ve outlined your business woes, it’s time to figure out which type of agent can best tackle them. Let’s take a look at three common flavours:

1. Collaborative AI agents
   Picture a small family of AI agents all pitching in to complete a task. AirOps is a great example: it’s a “content orchestration system” that taps into multiple tools and strategies to produce top-notch, SEO-friendly content—overseen by a real human to ensure quality. It’s like having a mini marketing team working 24/7!
2. Automation AI agents
   These whizz-kids can handle entire tasks and processes with minimal (or sometimes zero) human input. Take Otter Pilot from Otter.ai: it automatically hops into virtual meetings, records and transcribes them, then fires off a tidy summary and action items to Slack or email. Essentially, it’s your personal meeting scribe—but one that’s never late, never tunes out, and never complains about taking notes.
3. Social AI agents
   More people-focused, these are the chatty types. They excel at customer support, appointment scheduling, and giving you tailored information without forcing you to scour the web. If you’re dreaming of a kid-friendly, all-inclusive beach holiday that’s within driving distance and under a strict budget, a social AI agent can serve up your perfect itinerary—no more sifting through pages of reviews or questionable travel blogs.

Building And Releasing Your AI Agent

Now, if you’re a non-techie founder and the thought of building an AI agent makes you break out in a cold sweat, don’t worry. There are oodles of no-code resources out there that have your back. According to a white paper recently released by Google, two standout platforms are LangChain and Vertex AI.

LangChain, an open-source framework, is especially handy for connecting Large Language Models (LLMs) to external data sources. Meanwhile, Vertex AI lets you train, deploy, and customise AI models and applications in a snap—perfect for development teams that want to focus on finessing their agents instead of juggling the complexities of model building. You can read more about LangChain here, and visit the Vertex AI Studio here.

However you choose to build your AI agent, make sure you don’t leap straight into a massive, enterprise-wide rollout. Take it step-by-step. Begin with a small pilot, gather feedback, spot any bugs or bizarre behaviours, and fix them before letting your new digital assistant run riot across your entire organisation.

Why bother with this slow-and-steady approach? Well, as Google’s white paper notes, “no two agents are created alike due to the generative nature of the foundational models that underpin their architecture” 2023 Google White Paper on Generative AI. In plain English: your AI agent is going to behave uniquely, and the only way to refine it is through experimentation, feedback, and ongoing tweaking. That’s how you’ll strike gold (or at least avoid a meltdown).

The Human Touch Remains Essential

With all the hype and glittering potential of AI agents, it’s tempting to think: Set it and forget it. But let’s pump the brakes there.

In other words, your AI agent is not a magic wand. Yes, it can supercharge productivity and slash tedious busywork, but it still needs oversight, guidance, and purpose. Think of AI agents like members of your team: train them, guide them, set the right objectives, and you’ll unlock dazzling new levels of efficiency and creativity. Let them run rampant without proper guardrails, and, well, don’t be surprised if something goes awry.

Final Thoughts

AI agents have already begun to reshape the global business landscape. From automating everyday tasks to orchestrating more complex, multi-step projects, they’ve become indispensable for companies looking to stay ahead. But success depends on identifying why you need them in the first place, choosing the right one for the job, rolling it out carefully, and giving it ongoing human supervision.

Will 2025 truly be the year of the AI agent? If current trends are anything to go by, absolutely. Businesses that embrace AI agents with strategy, foresight, and a healthy dose of realism stand to gain a competitive edge in the coming years. Will you be one of them?

You may also like:

• Microsoft’s AI Agents Set to Transform Asian Workplaces
• How Digital Agents Will Transform the Future of Work
• Unleashing the Power of AI Agents

TL;DR – What You Need to Know in 30 Seconds

• Digital transformation often ended up as digitising old processes rather than fundamentally reinventing them.
• Research from KPMG shows 51% of companies haven’t seen performance gains from digital investments, and Gartner notes only 19% of boards are making real digital progress.
• AI-first transformation means using AI to connect all departments, data sources, and workflows into a single intelligent enterprise.
• Siloed thinking is no longer viable. AI thrives on cross-functional data and collaboration.
• AI-first companies have the chance to become the new Amazons and Ubers of the world, delivering exponential—rather than incremental—value.A truly AI-first system is more than a tool; it’s an enterprise-wide OS that learns, automates, and augments tasks and decisions in real time.
• The potential for Large Action Models (LAMs) suggests that AI could soon be doing far more than assisting with tasks—it could be acting on your behalf across the enterprise.

AI-first Business Transformation—Wht You Need to Know

Let’s have a little chat about something that’s been on the minds of everyone from the boardroom to the breakroom: transformation. But I’m not talking about the usual “digital transformation” we’ve all been hearing about for yonks. I’m talking about the next big wave that’s crashing onto our shores: AI-first business transformation. You might be thinking: “Haven’t we done this dance already? We’ve invested in digital, we’ve got shiny new software packages, we’re in the cloud… we’re modern, right?” Well, not exactly.

In truth, many of us may have got “digital transformation” a bit muddled. Rather than truly transforming our organisations, we seemed to simply digitalise them, upgrading existing processes instead of tearing them down and reimagining them from the ground up. Luckily, the rise of artificial intelligence is giving us that second shot at greatness—an opportunity to do more than just make existing models faster or better. Instead, AI lets us tackle an entirely new way of doing business, fundamentally rethinking how our enterprises operate, how our people collaborate, and how we measure success in a rapidly changing world.

Now, let’s roll up our sleeves and explore what it truly means to move from your “digital transformation” checklists to an AI-first mindset—and why this time around, we’ll actually transform.

We (Sort Of) Missed the Mark with Digital Transformation

Think about the promises made in the early days of digital transformation. We were told that new technologies would help us reinvent how businesses run. There’d be synergy, new models, dynamic reinvention of processes, cross-functional collaboration, you name it. Yet, if you look closely at what happened, we mostly digitised what we already did:

• Traditional processes got a digital facelift.
• Departments introduced new software, but largely worked the same way as before.
• Legacy mindsets remained intact, albeit with new jargon.
• Data continued to live in siloed systems designed for each function.

The result? We invested loads of money in “digital transformation” without always seeing the returns we were promised. Here’s a tidbit to put this in perspective: KPMG research reveals that 51% of companies have not seen an increase in performance or profitability from digital investments. That’s a majority who haven’t reaped the anticipated rewards. Equally sobering, Gartner found that only 19% of boards reported making progress toward achieving digital transformation goals. That’s not exactly the stuff of glowing quarterly reports, is it?

If you’re nodding in agreement (or maybe sighing in relief that you’re not the only one in this boat), you’re in good company. It seems many of us got stuck on the “digital” bit—throwing systems at old ways of working—rather than delivering true “transformation.” We reformed our businesses with technology. What we didn’t do was fundamentally reimagine them for a truly digital-first (and now AI-first) world.

Digital-First Companies Showed Us Another Way

While the majority of organisations were busy digitally updating and reformatting, a few outliers emerged, totally rethinking how their businesses should operate. Enter your classic “born-digital” or “digital-first” players:

• Amazon: Didn’t just sell books online; they transformed the entire commerce landscape to put digital at the heart of the retail experience.
• Netflix: Moved from DVD mail-outs to streaming, rethinking the very notion of consuming entertainment.
• Uber: Turned the transportation industry on its head with an on-demand, digital-first model.
• Airbnb: Revolutionised the hospitality sector without owning a single hotel.
• Twitch: Reinvented gaming by pairing it with social interactivity and live streaming.
• DoorDash: Did for delivery what Amazon did for retail, creating convenience and instant fulfilment that simply wasn’t possible in the old models.

These digital-first businesses didn’t just lob a piece of software at existing structures; they fundamentally re-engineered how those industries operated. The lesson here? If you’re going to embrace new tech, you have to also challenge conventional ways of thinking. Amazon didn’t just add a website to a bookstore; Netflix didn’t merely digitise DVDs. They scrapped legacy processes, mindsets, and assumptions—and came out on top because of it.

Now, with artificial intelligence (AI) shaking up the playing field in ways we’ve never seen before, we have a new chance to become “AI-first” enterprises—if we learn from the mistakes of the digital transformation era.

Digital Transformation Was the “How,” AI Is the “Why”

Digital transformation improved the way we do things, but often stayed stuck in departmental silos:

• HR had Workday.
• Sales had Salesforce.
• Marketing had HubSpot or Adobe solutions.
• Finance and supply chain had SAP.

But rarely did we ask: Should these processes continue to exist as they are, or could we re-engineer them completely? Instead, each group plugged in its digital solution, rarely integrating them into an overarching business framework. That, in turn, left data and workflows further fragmented, and sometimes it even added complexity.

Then along comes AI. AI doesn’t just give us a new tool; it promises a new paradigm. If used correctly, it compels us to connect the dots—across data, across workflows, across human resources, and ultimately across business units.

No more slicing and dicing by department. No more “We’ll just do the same old thing, only with AI to speed us up.” Instead, with an AI-first approach, we need to ask ourselves: How can AI help us see across the entire organisation to reimagine what’s possible? AI is the “why” we need to engage in a fundamental rethink of our operating model. Why keep HR, finance, marketing, and logistics so thoroughly compartmentalised? Why assume that the best way to manage your people, customers, and suppliers is with software that was effectively modelled after 20th-century workflows?

Here’s the rub with silos: work doesn’t stay in silos. Tasks and data typically move from one department to another. If you isolate improvements within a single department, you’re leaving enormous amounts of potential synergy untapped. Picture an ultra-optimised marketing CRM that can handle leads like a dream—but the supply chain can’t keep pace, the sales team has no cross-function visibility, and customer service is clueless about the marketing pipeline. You can guess how well that serves the customer or the bottom line.

We can’t just let each department run off and build its own AI tool. That might create pockets of brilliance, but it stops short of true transformation. Instead, it’s time for us to start imagining a connected enterprise that uses AI to flow insights and decisions throughout the entire organisation in real time. If your AI in customer service identifies a new product usage trend, that insight should feed into marketing, product design, logistics, you name it. Think of AI as the ultimate traffic controller: it routes the right data to the right place at the right time, helping you make sharper decisions that serve the greater good.

But let’s be clear: achieving that level of interconnectedness isn’t as simple as flipping a switch. We need new ways of structuring our businesses, new forms of collaboration between different teams, and new ways of training our workforce to think beyond their departmental boundaries. That’s the kind of stuff that terrifies many leaders, but if we’re serious about AI-first business transformation, it’s precisely where we have to go.

Shifting to an AI-First Mindset

An AI-first mindset says that if you have an HR workflow, for example, you don’t just ask how to automate or expedite it. Instead, you step back and ask: Is there an entirely new way to handle HR in the age of AI? Rather than just letting HR live in its digital system, can we integrate HR processes with other workflows—like IT provisioning, project management, or performance reviews—so that employees and managers see a single, seamless interface for all their needs?

In reality, you’ll find that no one’s job is as isolated as it might appear on an org chart. An HR leader also sits on cross-functional committees. A marketing person may weigh in on product design. A finance person is also an internal user of the IT helpdesk. When we isolate everything, we wind up making these cross-functional tasks painfully convoluted. An AI-first approach can do more than connect the dots; it can predict the best route through the entire enterprise, bridging these work streams effortlessly.

Not to be overly dramatic, but if you can harness AI to link these processes end to end, your daily workflows become the launching pad for a whole new level of productivity. No more duplicative data entry, no more emailing spreadsheets or chasing sign-offs in multiple systems. Instead, you’ll have an enterprise-wide “plumbing” that’s constantly learning and optimising itself so that the next time a similar task arises, you can handle it in half the time with half the fuss.

The Augmented Enterprise: When Everything Connects

So what does an AI-first business transformation look like in action? Once you break down silos and let AI do its thing, you get what some call the augmented enterprise. Essentially, AI augments:

1. Your People: Employees are guided by AI insights, making them more efficient and creative in their roles. Repetitive tasks can be automated or partially handled by AI agents, freeing people to focus on innovation and strategic thinking.
2. Your Processes: Workflows are streamlined and connected across departments. AI not only speeds them up but also surfaces predictive insights, letting you solve issues before they snowball.
3. Your Data: No more data living in locked compartments. An AI-first approach unifies data so that it can be analysed holistically—ensuring you spot patterns that were previously invisible.

Eventually, we might even see Large Language Models (LLMs) morph into something like Large Action Models (LAMs)—where AI doesn’t just summarise text or produce content, but actually takes actions on your behalf, in line with the business’s strategic goals. That’s an entire shift to AI as agent rather than AI as tool.

Is it futuristic? Sure. But it’s closer than you think. The more we interconnect these systems, the more potential there is for AI to genuinely run certain processes autonomously, or at least semi-autonomously. And that’s where transformation stops being linear and becomes exponential.

AI-First Companies: The Next Generation

If you missed the boat on being “digital-first,” don’t fret. Right now, there’s an opportunity to be among the first wave of “AI-first” organisations. The possibilities are massive:

• Product Development: AI can shorten product lifecycles by analysing performance data, testing new features through simulation, and even generating prototypes.
• Customer Experience: AI can unify your CRM, chatbots, and call centre workflows, ensuring you respond to queries with instant knowledge of the customer’s history, preferences, and future needs.
• Supply Chain Management: AI can predict demand surges, optimise shipping routes, and even manage inventory in real time, preventing bottlenecks that cost you money and customers.
• Finance & Accounting: Automated processes for invoicing, expense management, and forecasting. Your finance team becomes data-driven analysts, leaving behind laborious manual tasks.
• Human Resources: AI can screen applicants, highlight training needs, and pinpoint cultural or engagement issues before they turn into full-blown crises.

In a few years, people might talk about the big AI-first successes the same way they talk about Amazon or Netflix today. If you seize the day, your company could be among them. As Sam Altman, CEO of OpenAI, quipped: “This is the most interesting year in human history, except for all future years.” That’s quite the statement—and it’s a reminder that we’re just at the beginning of what AI can accomplish.

Designing an Intelligent Enterprise Operating System

Picture an enterprise-wide system of intelligence—a single, integrated platform that links up every function, every data source, and every person:

1. Unified Data Layer: All your data from across departments is fed into a single AI backbone. This is crucial because AI needs vast, high-quality datasets to produce its best insights.
2. AI Agents Everywhere: Intelligent virtual assistants embedded in each department, not just to execute tasks but to interpret them, predict outcomes, and suggest next steps.
3. Cross-Functional Collaboration by Design: No more departmental silos because your system fundamentally disallows them. Project creation, resource allocation, and approvals all happen within the same architecture, with AI facilitating smooth transitions.
4. Continuous Improvement: As the system runs, it gathers more data about how tasks are accomplished and outcomes achieved. AI uses this to refine its own recommendations—compounding your improvements exponentially rather than linearly.
5. Focus on Innovation: When day-to-day tasks get automated or augmented, you free human capital. These employees can then channel their creativity into new revenue streams, product ideas, or strategic initiatives.

That, my friends, is what an AI-first business transformation boils down to: not merely accelerating old processes but reimagining the entire way you do business, top to bottom, front to back.

The Future Is Exponential

We’re standing on the cusp of a new era, one where “digital transformation” might look like a quaint stepping stone. AI has the potential to create the sort of exponential leaps that 20th-century businesses could only dream of. If we seize the opportunity, we won’t just see incremental gains; we’ll witness leaps in productivity, the birth of entirely new business models, and a surge in personalised, data-driven solutions that deliver value for everyone—customers, employees, and stakeholders alike.

As the world keeps shifting under our feet, one thing remains crystal clear: standing still is not an option. Doing nothing, or doing the bare minimum, risks being left behind by those who adopt AI-first strategies early and wholeheartedly. And if we learned anything from the digital-first revolution, it’s that latecomers can catch up, but it’s a much harder road.

So, here’s your rallying cry: challenge every assumption, connect every silo, unify every dataset, and bring in AI not just as another tool, but as a co-creator of your future enterprise. The age of linear growth and departmental thinking is drawing to a close. The time of interconnected, exponentially enabled businesses is here.

What Do YOU Think?

Will you settle for being a digital dinosaur stuck in the old ways, or will you harness AI to boldly redefine your organisation for a future where transformation is continuous, interconnected, and exponentially powerful? Let us know in the comments below!

Let’s Talk AI!

How are you preparing for the AI-driven future? What questions are you training yourself to ask? Drop your thoughts in the comments, share this with your network, and subscribe for more deep dives into AI’s impact on work, life, and everything in between.

You may also like:

• AI Integration Headache: Connecting the Dots in Asia’s Tech Boom
• Adrian’s Arena: Reaching Today’s Consumers – How AI Enhances Digital Marketing
• ChatGPT Canvas: The Future of AI Collaboration is Here!
• Or try DeepSeek AI by tapping here.

TL;DR – What You Need to Know in 30 Seconds

• Perplexity’s Deep Research tool offers advanced AI research capabilities for a fraction of typical enterprise costs.
• It provides five free queries daily and charges $20 per month for 500 queries—compared to big AI providers charging thousands.
• Scored 93.9% on SimpleQA and 20.5% on Humanity’s Last Exam, outpacing Google’s Gemini Thinking, with OpenAI only slightly higher at 26.6%.
• Enterprise AI spending is projected to rise by 5.7% in 2025, although some companies are increasing their AI budget by 10% or more.Deep Research could shift the market by making companies question premium AI subscriptions that cost up to 100x more.
• The tool handles a range of tasks (healthcare, finance, market research) in under three minutes, democratising AI for smaller businesses and individuals.
• This affirms a new era in AI, where affordability meets performance, and big spenders must now justify their exorbitant costs.

Unpacking Perplexity Deep Research Tool, and its Impact

Today, we’re diving into one of the most talked-about innovations in AI right now: Perplexity’s new Deep Research tool. If you haven’t heard of it yet, don’t fret—this is precisely what we’re here for. Grab your favourite cuppa, because we’re about to explore how Perplexity is turning AI research upside down, smashing cost barriers, and making us question every pricey AI subscription that’s ever crossed our desks. Sound good? Let’s get stuck in!

The Big Bang of Affordable AI

You know how some products come along and make you wonder why you ever paid so much for something else? That’s exactly what’s happening with Perplexity’s Deep Research. In a single, bold move, Perplexity has basically told the rest of the AI industry: “We’re here, we’re cheap, and we’re not messing about.” If you haven’t caught wind of it, Deep Research is a tool that can generate comprehensive research reports in just minutes. Yes, minutes. And here’s the kicker: it offers advanced AI capabilities at a fraction of the typical enterprise costs.

Take a look at what’s on the table: while Anthropic and OpenAI can easily charge into the thousands every month for their premium services, Perplexity is throwing in five free queries daily for all users and an upgrade at $20 per month for 500 daily queries plus faster processing speeds. That’s not just cheaper; it’s borderline scandalous when you see that other AI giants charge almost 100 times more for near-similar (and, in some cases, arguably lesser) capabilities.

But it’s not just a marketing gimmick. Aravind Srinivas, Perplexity’s CEO, shared the company’s ethos on X (formerly Twitter), saying, “Knowledge should be universally accessible and useful. Not kept behind obscenely expensive subscription plans that benefit the corporates, not in the interests of humanity!” It’s hard not to be inspired by that. The democratisation of AI has long been touted as the Next Big Thing in tech, but Perplexity is making some serious strides to actually achieve it, rather than just talk about it.

Deep Research is now a commodity thanks to Perplexity pic.twitter.com/Fk8yvPTLzV— Aravind Srinivas (@AravSrinivas) February 14, 2025

Enterprise AI Spending Under the Microscope

As you might guess, this sudden plunge in price is raising eyebrows—big time. Large enterprises have been funnelling massive budgets into AI, with some expecting to increase their AI spending by 5.7% in 2025. That’s despite overall IT budgets going up by less than 2%. In certain sectors, that surge in AI spending could be as high as 10%, and on average, some businesses plan to throw in an additional $3.4 million into AI initiatives. With the rise of Deep Research, though, those expensive subscriptions now look a little, well, questionable.

Let’s be real. When you’ve got a brand-new AI tool that gives near-enterprise level performance (and sometimes even more advanced capabilities) for $20 a month, it begs the question: What are we actually paying for with those premium AI subscriptions? If you’re on the corporate side, you might be reviewing your budgets as we speak. Think about the training, the data hosting, the staff overhead—yes, those are real costs. But are they enough to justify a 100x difference in price?

Technical Mastery That’s Giving Giants a Run for Their Money

Now, let’s talk numbers, because who doesn’t love a good metric? Perplexity’s Deep Research scored a whopping 93.9% accuracy on the SimpleQA benchmark and clocked 20.5% on Humanity’s Last Exam. If you’re wondering why that second number is interesting, consider that it outperforms Google’s Gemini Thinking and other top-tier models. Even more eyebrow-raising is that OpenAI scores 26.6% on Humanity’s Last Exam—yes, that’s higher than Perplexity’s 20.5%—but let’s not forget the monstrous cost difference for that extra 6 percentage points.

Perplexity also claims that Deep Research completes most tasks in under three minutes, performing dozens of searches and analysing hundreds of sources simultaneously. That’s lightning-fast by any measure, especially when you realise it’s essentially replicating what expert human researchers would do—but in a fraction of the time. For advanced tasks like financial analysis, market research, technical documentation, or even healthcare insights, it’s an absolute game-changer.

Why This Matters to You (and Everyone Else)

Alright, it’s cheap, it’s fast, and it’s accurate. Who cares, right? Well, pretty much anyone who’s ever wanted to make use of advanced AI capabilities but balked at the price tag. It’s no secret that enterprise AI has often ended up creating a digital divide: if you’ve got the budget, you can do some serious data-crunching, but if not, you’re left in the dark ages. This means smaller businesses, individual researchers, students, or freelancers could only dream of some of these AI services because they couldn’t justify the cost.

But along comes Perplexity, democratising the whole playing field. The potential is enormous. If you’re a small tech start-up, you no longer have to pay thousands just to get your data insights. Researchers can use Deep Research for thorough academic or industry analyses. Professionals in healthcare or finance can produce detailed reports that would usually cost an arm and a leg. And because Perplexity plans to expand Deep Research to iOS, Android, and Mac platforms, access is only going to get easier.

Is Enterprise AI in for a Shake-up?

If you’re in charge of procurement or strategic decisions for a big firm, your job just got a bit more complicated. Do you stick with the big-name provider with that hefty subscription fee, or do you try Perplexity to see if it meets your organisation’s needs? The key question is: Are you really getting the added value for your money when your monthly AI bills are in the thousands?

Sure, there could be a few reasons to keep paying extra. Perhaps you’re already deeply integrated with a certain AI ecosystem, or you need custom solutions that only a big player can provide. Maybe you rely on dedicated customer support that’s included with your pricy subscription. But the argument that premium cost automatically translates to premium capability is quickly losing steam.

With Perplexity’s impressive performance, we might see a future where expensive enterprise AI tools have to scramble to prove they’re worth it. You can’t just plaster “enterprise-grade” on a service and watch the money roll in—users want tangible, cost-effective results.

How Deep Research Outperforms (and Where It May Still Lag)

Let’s not gloss over the fact that OpenAI’s own research capabilities still technically inch out ahead in certain benchmarks. A 26.6% score on Humanity’s Last Exam compared to Perplexity’s 20.5% might be a big deal for mission-critical tasks in specialised domains. Then again, Perplexity’s 93.9% on SimpleQA is hardly peanuts. And let’s remember the price difference—OpenAI can charge hundreds (if not thousands) of percent more. So is that extra 6 percentage points in performance worth the colossal increase in cost?

It all boils down to your use case. If you’re a hedge fund manager who needs the absolute best of the best and every fraction of a percent could mean millions in revenue, you might still throw your money at the top-of-the-line model. But if you’re a mid-sized firm or an independent researcher, Perplexity’s offering is more than enough—especially at $20 a month.

Practical Implications: From Healthcare to Finance

Let’s look at some real-world scenarios. Healthcare professionals can use Deep Research to scour medical journals, clinical trial results, and official guidelines faster than you can say “NHS queue”. This means better patient outcomes, quicker insights, and less reliance on massive IT budgets.

Financial analysts can crunch market data, follow the latest economic news, and whip up in-depth reports that previously needed entire teams of well-paid data scientists. Technical documentation tasks become a breeze when Deep Research can parse through troves of manuals, development forums, and official documents in minutes.

Plus, Perplexity’s user-friendly features—like exporting findings as PDFs or sharing them directly through its platform—make collaboration straightforward. If you’ve ever had to wrestle with clunky enterprise software, you’ll appreciate the simplicity that Perplexity offers.

The Democratisation Ripple Effect

We’ve talked about how smaller entities stand to benefit from cheaper AI tools. But let’s not forget the social dimension. When you lower the barrier to entry, you empower not just businesses, but also students, civil society organisations, journalists, and independent researchers. Knowledge stops being locked behind corporate walls. That’s a big deal in Asia—where the digital transformation wave is sweeping nations at very different speeds and scales.

Imagine an NGO in a rural part of Southeast Asia that can now access top-notch AI research capabilities for $20 a month. That’s a giant leap forward in bridging the digital gap, enabling them to better serve local communities, gather data, and deliver more effective programmes. It’s not just a business story; it’s a social justice story too.

What’s Next in 2025 and Beyond?

Given that AI spending is expected to rise by 5.7% in 2025, the question on everyone’s lips is how this new wave of budget-friendly AI offerings will redistribute the market. Will companies continue throwing millions at established AI giants, or will they pivot to nimble, cost-effective alternatives like Perplexity?

In many ways, this sets the stage for an AI arms race of affordability and performance, where large players need to prove they’re worth the extra cash—or risk losing market share. From what we’re seeing, the AI community (and the public) are hungry for an open-source, reasonably priced alternative. Perplexity’s decision to offer a free daily query allowance and then a generous 500 queries a day for a mere $20 might be the blueprint for the future of AI subscription models.

The Jury’s Verdict and a New Era

So, does this mark the end of expensive AI subscriptions? We’ll have to wait and see. But one thing is clear: Perplexity’s Deep Research has seriously called into question the notion that you need to pay through the nose for quality AI. If performance is almost on par with more expensive services, or in some benchmarks better, why wouldn’t you jump on board?

Perhaps the biggest indicator of success will be user adoption. And it’s already looking promising—thousands of folks have begun testing Deep Research, singing its praises, and pondering whether they really need those thousand-dollar monthly fees. In the dynamic, ever-shifting AI world, the best technology won’t be the one with the biggest marketing budget but the one that’s truly accessible to the people who need it most.

There you have it, folks: a whirlwind tour of how a single innovation from Perplexity is rattling the foundations of AI’s business model. Whether you’re a budding researcher, a startup founder, or a corporate decision-maker, the paradigm is changing right before your eyes. Will you be part of the revolution—or left clinging to yesterday’s overpriced subscriptions? The choice, as always, is yours! And don’t forget to subscribe to keep up to date with all the latest AI happenings in Asia.

What Do YOU Think?

As AI costs plummet and quality soars, will businesses continue to pay premium prices out of habit—or dare to embrace the affordable future? Let us know in the comments below!

You may also like:

• Perplexity Deep Research Tool Debuts, Challenging OpenAI and Google
• Perplexity Assistant: The New AI Contender Taking on ChatGPT and Gemini
• Or try Perplexity AI by tapping here.