How SEA’s edtech firms are taking on ChatGPT

In The Prompt this week, we analyze the threat ChatGPT poses to SEA’s edtech startups and the high costs stalling companies from genAI adoption. [Read from your browser]( The Prompt 🤖 Welcome to the second edition of The Prompt! Delivered every fortnight, this free newsletter breaks down the biggest stories and trends in AI. Get all our content, including The Prompt, by [registering for a Tech in Asia account](. Alternatively, [sign up here]( if you just want to receive this newsletter. IN FOCUS In today's newsletter, we look at: - [ChatGPT threatens edtech startups, but founders are unfazed]( - [Why Indonesia trails in AI investment in Southeast Asia]( - High compute costs of genAI brings ROI into question --------------------------------------------------------------- Hello {NAME} The divide between reality and fiction seems to diminish day by day. I find it remarkable that humans are able to turn science fiction ideas into reality, whether it’s venturing into space, creating flying cars, and more recently, developing artificial intelligence that gives machines human-like speech. In May, OpenAI unveiled its voice assistant product GPT-4o, with co-founder and CEO Sam Altman referring to the fictional AI Sky in the [movie Her](. It seemed clear to me where the inspiration for GPT-4o came from. While AI voice assistants such as Siri and Alexa have been around for a while, what captivated everyone’s attention was GPT-4o’s ability to display human-like emotions through the tone of voice. One of the applications OpenAI showcased for GPT-4o was for [the edtech sector]( where the AI can teach a child various concepts. This raises the issue of whether edtech startups that have developed their own AI voice assistants might be disrupted. This brings us to our Big Story for today, where my colleague Elyssa speaks to edtech founders in Southeast Asia to find out how big a threat ChatGPT poses to their products. Some founders remain unfazed about the generative AI giant’s capabilities, as they believe they offer a differentiated product. But other players such as CoLearn aren’t jumping to create new AI-enabled services, as a co-founder says it’s “not cheap to run an AI model.” Meanwhile, in the Hot Take, I look at a contrarian view that genAI isn’t reducing costs for companies but, in fact, increasing them. Last week, a Gartner analyst stated that 30% of genAI projects will be abandoned at the proof-of-concept (PoC) stage by the end of 2025, mainly due to significant compute costs. Already, global IT companies working on genAI projects for clients have stalled from scaling their PoCs to products that can go out to customers. I take a closer look at how the cost factor might hinder genAI development. But first, the Big Story! -- Collin  --------------------------------------------------------------- THE BIG STORY [ChatGPT threatens edtech startups, but founders are unfazed]( Personalized services and the accuracy of AI-enabled models will be the next competitive areas for Southeast Asia's edtech startups.  --------------------------------------------------------------- THE HOT TAKE Companies are quickly learning that genAI isn’t pocket friendly Here’s what happened: - By end-2025, 30% of genAI projects will be abandoned at the proof-of-concept (PoC) stage, as predicted by market research firm Gartner. - Compute costs have been rising significantly with growing demand for cloud GPUs and high energy requirements. - Some experts predict that instead of driving cost savings, companies using AI will see a significant increase in their expenses over the years. Here’s our take: One of the hottest topics being debated recently is the soaring compute costs of generative AI. It is a major reason why many genAI projects are getting delayed. Last week, global market research firm Gartner [predicted]( that by the end of 2025, at least 30% of genAI projects will be abandoned after the PoC stage due to escalating costs, uncertain business value, poor quality of data, and insufficient risk controls. While many companies and startups today want to use genAI to transform their business models and create new revenue streams, the cost to do so is significant, ranging from US$5 million to US$20 million. The question for many CFOs is whether the scaling of PoCs into actual products is worth the return on investment. Already, many Indian IT companies are facing [significant delays]( in moving their PoCs for clients to the production-grade product stage. This is primarily due to the high compute costs, which can be attributed to the hyperscalers - cloud GPU and storage service providers such as AWS, Microsoft Azure, and Google Cloud - having high-price models. These hyperscalers are offering their latest and most accurate foundational models for clients that are ready to make annual commitments. But those who opt for the pay-as-you-go option may only get an older model. The latter means dealing with accuracy issues, as [AI hallucinations]( are more prevalent in older models, leading to increased costs down the line. Currently, OpenAI’s GPT-4 models [cost US$0.06 per 1,000 tokens]( - each token represents the smallest unit into which the data can be broken down for the AI model to process - which is 30x more expensive than GPT-3.5 models. Though the price tag might seem small, it adds up as a user would utilize 100 tokens for a prompt of around 75 words. In addition to the costs, finding good quality comprehensive data to train these genAI products is limited, and they're only as good as the data they're trained with. These factors have led to companies picking only a few of their PoCs to scale to production-grade products. This is why many believe that the costs of companies integrating genAI into their products will increase rather than go down. While some investors like Vinod Khosla, founder of Khosla Ventures, remain optimistic that [compute costs will decrease]( the math seems to show that this is highly unlikely, as [this founder illustrates](.  --------------------------------------------------------------- DEEP READS Check out these interesting reads about AI: 1️⃣ [Diamonds scarce in AI startup boom - but not impossible to find]( This article explores one of the burning questions in AI for investors: which startups will emerge as winners amid the barrage of genAI startups that exist today? It also asks whether we are seeing a bubble in the AI market, similar to the dot-com bubble in the late '90s. 2️⃣ [Is generative AI in drug discovery overhyped?]( In every industry, people are starting to look past the initial hype of genAI’s potential to revolutionize it. For instance, in the healthtech sector, genAI was expected to help speed up the drug discovery process. But this article’s author - who is the CEO of Insilico Medicine and has 10 years of experience in drug discovery - shares his experience about why the drug discovery process takes so long that even genAI can’t reduce the time by much. 3️⃣ [Generative AI development requires a different approach to testing]( Testing for coders is a very time-consuming process and can take days to complete. This article explores the use of genAI in helping developers test in a more efficient way. However, there are some drawbacks in the testing process when using the tech, and companies are looking at ways in which they can overcome them. --------------------------------------------------------------- SYNC IS BACK, AND THIS TIME, THE DIAL’S GOING TO ELEVEN Register your interest now to get the latest on all things #SYNC2024 Remember last year's SYNC? Tech in Asia handpicked 100 attendees for a day of learning, sharing, and networking, all aimed at addressing climate tech’s challenges and mapping the future of the sector. This year, we're focusing on another exciting industry, and we’re turning up the dial with even more thought-provoking sessions, inspiring speakers, and rising stars ready to disrupt. We'll bring together 100 of Asia's leading minds for an experience you won't want to miss. Ready to be in the know? [Register your interest today]( for the latest on SYNC 2024's official launch, agenda, pricing, and more. Stay tuned.  --------------------------------------------------------------- FYI [Why Indonesia trails in AI investment in Southeast Asia]( Since 2023, tech giants have committed US$1.9 billion for AI-related investments in Indonesia, which is only 3.7% of the total sum for the region.  --------------------------------------------------------------- EVENTS HAPPENING You can also check out a curated list of trending tech events [over here]( and Tech in Asia’s signature events [here](. [Tech in Asia Conference Jakarta on Oct 23 - 24]( Tech in Asia is at the forefront of uncovering the next big thing in Southeast Asia’s tech scene. Our flagship conference in Jakarta is where the region’s most promising trends converge. This year, we’re spotlighting Fintech, Cleantech, and AI – three sectors poised to revolutionize industries and create immense opportunities. Join us to gain invaluable insights and network with the region's top innovators. [Save 50% off your tickets today!]( [Startup Factory at Tech in Asia Conference Jakarta]( Are you building a startup that is on track to become the next tech frontier? Apply to Startup Arena, the hottest pitch battle judged by investors, industry experts, and a live audience. Vie for global recognition, prize money, media coverage, and other exciting prizes! [Apply now.](  [Gemini Day for Startups: Singapore on August 11]( The future of your AI-powered startup starts now. On August 22, get an exclusive preview of Google Cloud's latest AI tools and learn how to leverage their full potential. We're talking cutting-edge demos, unfiltered advice from AI veterans, and unparalleled networking that will make your LinkedIn profile a magnet for opportunity. [Apply now for an invite - limited seats up for grabs.]( [Tech in Asia’s Founders Meetup: Thailand on September 11]( Building a startup in Bangkok is a marathon, not a tuk-tuk race. Ready to accelerate your growth? Join us and connect with the city's top founders, leaders, and investors. Gain invaluable insights on navigating Bangkok's tech scene, and snag face-time with leading VCs during our exclusive Office Hours. Early-bird tickets are at a discounted price of US$10 until August 26. [Get your ticket now!]( --------------------------------------------------------------- That’s it for this edition - we hope you liked it! Not your cup of tea? You can unsubscribe from this newsletter [here](. In the meantime, if you have any feedback or ideas, feel free to get in touch with Terence, our editor-in-chief, at terence@techinasia.com. See you soon! P.S. Don't miss out on the biggest tech news and analysis. Add newsletter@techinasia.com to your address book, contacts, or safe sender list. Or simply move us into your inbox. Too many emails? Switch to a different frequency or get new content through our [preference center](. You can also break our hearts and remove yourself from all Tech in Asia emails over [here](  Copyright © 2024 Tech in Asia, All rights reserved. 63 Robinson Road, Singapore 068894