Newsletter Subject

Pizzas that come with a slice of AI and robotics

From

techinasia.com

Email Address

newsletter@techinasia.com

Sent On

Tue, Feb 27, 2024 02:00 AM

Email Preheader Text

This week’s On the Rise dives into an AI-powered pizza startup, pivots in the insect-for-food i

This week’s On the Rise dives into an AI-powered pizza startup, pivots in the insect-for-food industry, and the legal woes of AI hallucinations. [Read from your browser]( On the Rise 🚀 Welcome to On the Rise! Delivered every Tuesday via email and through the Tech in Asia website, this free newsletter breaks down the biggest stories and trends in emerging tech. If you’re not a subscriber, get access by [registering here](. --------------------------------------------------------------- IN FOCUS In today's newsletter, we look at: - A [pizza served]( using AI and robotics - The challenges of [raising insects]( - The growing conundrum of AI hallucinations Hello {NAME} , New York-style pizza tastes like tomato sauce smeared on cardboard. These are fighting words in some areas of the pizza loving world, but I stand by them. If I really want to start a ruckus with the pro-New York crowd, though, I always drop this bomb: I love pineapple on my pizza. A chicken pizza with feta cheese and cashews is also excellent. Despite the dominance of pizza snobs such as myself in online discussions, there is a third group of pie-loving people out there: the lunch crowd. They want something to eat, ideally quickly, and without argument about sauces. I may not agree with them, but I can appreciate their “I ordered lunch, not an argument” practicality. So with that group in mind, we have this week’s Big Story. My colleague Melissa talks to the founder and CEO of GoPizza, a South Korean chain that competes against fast food companies rather than against traditional pizza franchises. And yes, there is tech involved here. GoPizza uses AI to monitor the topping process, and robots add the condiments and sauces. After you’ve eaten - definitely not before - you should check out our second Big Story, which looks at the challenges of the insect-for-food industry, which has caused several startups to pivot. And then finally I'll tell you about how hallucinating computers wreaked havoc in the business world this week. -- Scott  --------------------------------------------------------------- THE BIG STORIES 1️⃣ [South Korean startup serves pizza with AI and robotics. But is it a flash in the pan?]( GoPizza, which uses AI and robots to serve up fast slices, has over 200 outlets in seven Asian markets and plans to open another hundred by June. 2️⃣ [Frustrations in insect-for-food industry spark pivot]( Some regional startups are finding better opportunities in producing insect-derived ingredients for health supplements and animal feed.  --------------------------------------------------------------- AI ODYSSEY Promising AI projects we’re noticing When genAI hallucinations court legal woes for companies The first time I encountered the concept of a hallucinating computer was in the pages of [Transmetropolitian]( a cyberpunk comic book from the late 1990s. In one of the first stories, antihero Spider Jerusalem has to deal with a hallucinating home appliance called a maker - basically a highly advanced 3D printer - which was making and consuming its own drugs. Of all the weird and outlandish concepts I’ve read in science-fiction over the years, digital hallucinations was not on my “things I’ll ever cover in real life” bingo card. Yet, a computer that was apparently off its rocker got [Air Canada]( hauled into court last week. An AI-controlled chatbot on the site gave erroneous and apparently made-up information - the plain English definition of an AI having a hallucination - to a traveler named Jake Moffat, who was seeking a bereavement fare. When the traveler followed the chatbot’s instructions, the airline refused to give a refund. Moffat sued, and Air Canada attempted to defend itself claiming the chatbot was a “separate legal entity that is responsible for its own actions.” The judge in the case issued a very common sense and non-technical ruling. Just because the information a customer relied on came from an AI-controlled chatbot and not plain text didn’t matter. Air Canada had put both sets of conflicting data on its website, and it wasn’t the customer’s duty to figure out which one was right. Air Canada owed Moffat the equivalent of US$812. The offending bot, which Air Canada said was only launched as an experiment, has since been pulled from the website. The bigger question is whether the case is an anomaly or part of a bigger problem with AI. Just how much hallucinating is going on out there? A study that came out just last month from [Stanford University]( seems to suggest that hallucination rates can sometimes reach horrific levels. The study, written by [Stanford RegLab]( and the [Institute for Human-Centered AI]( said that hallucination rates for “specific legal queries for state-of-the-art language models” range from 69% to 88%. While the authors of the Stanford paper make it clear that their research is limited to legal issues, AI-startup [Vectara]( has compiled a [leaderboard]( of the hallucination rates for general AI programs. ChatGPT-4 comes in with a 3% hallucination rate, the best ranking on the list. Google’s [Flan-T5]( pulls up the rear with a 15.8% rate. Why is there a huge difference between these rates and those reported by Stanford? The study found that an AI’s performance deteriorates when dealing with more complex tasks that require a “nuanced understanding” of the issues. Building an LLM that understands these nuanced relationships in complex systems is something that genAI companies will have to work toward. But the more immediate question for the AI industry and businesses in general is one we’d never have to ask if real people were involved: What is the acceptable level of hallucinations? As far as I can find, no one has been able to draw a line between hallucination rates and monetary losses. Unfortunately, that data will probably only come after more cases of genAI-created financial damage. -- Scott  --------------------------------------------------------------- FYI 1️⃣ [Three Arrows’ liquidators to pay out interim dividend in March]( The liquidators of Three Arrows Capital have announced they intend to distribute an interim dividend on or around Mar 31, 2024. 2️⃣ [A list of SEA’s unicorns and their early investors (updated)]( Silicon Box reaching billion-dollar status brings a glimmer of hope after Southeast Asia’s unicorn count dwindled from 2022 to 2023.  --------------------------------------------------------------- NEWS YOU SHOULD KNOW Also check out Tech in Asia’s coverage of the emerging tech scene [here]( 1️⃣ [Chinese genAI startup gets backing from Alibaba]( Moonshot AI has secured over US$1 billion in funding spearheaded by Alibaba and Chinese VC firm HongShan. 2️⃣ [Nvidia achieves record revenue]( The chips manufacturer reported a record full-year revenue of US$60.9 billion, a 126% increase from FYE 2023. 3️⃣ [Franklin Templeton bets on crypto news]( Asset manager Franklin Templeton has invested an undisclosed amount in Blockhead, a Singapore-based crypto news firm. The investment will support the launch of BRN, Blockhead’s new research platform, among other initiatives. 4️⃣ [VinFast to make EVs in India]( The Vietnam-based electric vehicle manufacturer broke ground on its new India facility on February 25. The company has committed to investing US$500 million for the first phase of the integrated EV factory. 5️⃣ [Byju’s backers battle founder]( The troubled edtech giant is facing a lawsuit filed by its own backers. Four major investors have filed a petition with India’s company tribunal seeking to remove founder Byju Raveendran and his wife, Divya Gokulnath, from leadership positions in the firm due to alleged oppression and mismanagement.  --------------------------------------------------------------- That’s it for this edition - we hope you liked it! Do also check out previous issues of the newsletter [here](. Not your cup of tea? You can unsubscribe from this newsletter by going to your “edit profile” page and choosing that option in our preference center. See you soon! [ADVERTISE]( | [SUBSCRIBE]( | [HIRE]( | [FIND JOBS]( 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]( or [unsubscribe](. 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

Marketing emails from techinasia.com

View More
Sent On

08/12/2024

Sent On

06/12/2024

Sent On

09/11/2024

Sent On

03/11/2024

Sent On

30/10/2024

Sent On

28/10/2024

Email Content Statistics

Subscribe Now

Subject Line Length

Data shows that subject lines with 6 to 10 words generated 21 percent higher open rate.

Subscribe Now

Average in this category

Subscribe Now

Number of Words

The more words in the content, the more time the user will need to spend reading. Get straight to the point with catchy short phrases and interesting photos and graphics.

Subscribe Now

Average in this category

Subscribe Now

Number of Images

More images or large images might cause the email to load slower. Aim for a balance of words and images.

Subscribe Now

Average in this category

Subscribe Now

Time to Read

Longer reading time requires more attention and patience from users. Aim for short phrases and catchy keywords.

Subscribe Now

Average in this category

Subscribe Now

Predicted open rate

Subscribe Now

Spam Score

Spam score is determined by a large number of checks performed on the content of the email. For the best delivery results, it is advised to lower your spam score as much as possible.

Subscribe Now

Flesch reading score

Flesch reading score measures how complex a text is. The lower the score, the more difficult the text is to read. The Flesch readability score uses the average length of your sentences (measured by the number of words) and the average number of syllables per word in an equation to calculate the reading ease. Text with a very high Flesch reading ease score (about 100) is straightforward and easy to read, with short sentences and no words of more than two syllables. Usually, a reading ease score of 60-70 is considered acceptable/normal for web copy.

Subscribe Now

Technologies

What powers this email? Every email we receive is parsed to determine the sending ESP and any additional email technologies used.

Subscribe Now

Email Size (not include images)

Font Used

No. Font Name
Subscribe Now

Copyright © 2019–2025 SimilarMail.