Tesla shares fall after production and deliveries lag due to logistic hurdles

By

-

October 4, 2022

Tesla shares fall after production and deliveries lag

Tesla’s shares fell 7% on Monday after the company sold fewer-than-expected vehicles in quarter three as deliveries lagged way behind production because of logistic hurdles.

The shares opened at a more than two-month low, and if losses hold, Tesla will lose about $58 billion in market capitalization in a single day. It will be more than the market value of General Motors Co. and Ford Motor Co.

Tesla produced a record 365,923 EVs in the quarter and delivered a new high of 343,830, but still missed market expectations, with a huge gap between production and deliveries raising investor worries.

The company said it could not secure enough transportation during the peak time at a reasonable cost for vehicles made. Chief Executive Elon Musk has vouched for steadier deliveries in the current quarter to reduce the last-minute rush.

Due to the increased prices of Tesla vehicles, higher borrowing costs, and a dull outlook for global economic growth, the shortfall in deliveries comes with demand worries among investors and analysts.

Tesla will need to deliver more than 450,000 vehicles in the fourth quarter to meet its goal of growing deliveries by 50% annually. The company aims to produce about 495,000 Model Y and Model 3 during the period. Shares of smaller electric-vehicle makers, including Rivian Automotive Inc, Fisker Inc, and Lucid Group, were also down between 2.7% and 6.8% in early trading.

Kim Kardashian fined $1.26m for promoting cryptocurrency on Instagram

By

Sahil Pawar

-

October 4, 2022

Kim Kardashian fined $1.26m for promoting cryptocurrency

Kim Kardashian has agreed to pay $1.26m to settle charges from the United States Securities and Exchange Commission (SEC) that she promoted a cryptocurrency on Instagram without disclosing she had been paid $250,000 to do so. The post has now been removed.

The SEC said that the reality TV star has agreed to comply with its ongoing investigation. The SEC said Kim Kardashian failed to disclose that she was being paid to publish a post on her Instagram handle about EMAX tokens, a crypto asset security offered by EthereumMax.

Kardashian’s Instagram post contained a link to the EthereumMax website, which provided instructions for potential investors to purchase EMAX tokens. Gurbir Grewal, director of the SEC’s division of enforcement, said that the federal securities laws clearly state that any celebrity or other individual promoting a crypto asset security must disclose the source, amount, and nature of compensation they received in exchange for the promotion. Kardashian has agreed not to promote any crypto asset securities for three years.

Kim Kardashian’s lawyer said that she has fully cooperated with the SEC from the very beginning and remains willing to do whatever is expected of her to assist the SEC in this matter. Kardashian wanted to get the matter behind her to avoid any protracted dispute. Her agreement with the SEC allows her to do the same and move forward with her many business pursuits.

Cryptocurrency is facing increasing attention from the US Congress. The latest bipartisan proposal on crypto came in August from Senators Debbie Stabenow, a Democrat from Michigan, and John Boozman, a Republican from Arkansas. It would hand the regulatory authority over Bitcoin and Ether to the Commodities Futures Trading Commission.

Meta launches AI software tool AITemplate to switch between underlying chips

By

Sahil Pawar

-

October 4, 2022

Meta launches AI software tool AITemplate

Meta AI has developed and is open-sourcing AITemplate (AIT), a unified inference system with separate acceleration back-ends for both AMD and NVIDIA GPU hardware.

AITemplate delivers close to hardware-native Tensor Core (NVIDIA GPU) and Matrix Core (AMD GPU) performance on various widely used AI models such as convolutional neural networks, transformers, and diffusers. With AIT, it is now possible to run performant inference on hardware from both GPU providers.

Meta said its open-source AI platform is based on an open-source machine learning framework called PyTorch. According to Meta, AITemplate can help code run up to 12 times faster on Nvidia’s A100 chip or up to 4 times faster on Advanced Micro Devices (AMD) MI250 chip.

The software has become a battleground for chipmakers looking to build up an ecosystem of developers that utilize their chips. Nvidia’s CUDA platform continues to be the most popular so far for AI work.

Once developers tailor their code for Nvidia chips, running it on graphics processing units or GPUs from Nvidia competitors like AMD is not easy. AITemplate is designed to quickly swap between chips without being locked in, Meta said in a blog post.

Meta said that the GPU back-end support gives deep learning developers more hardware vendor choices with minimal migration costs. The company added that proprietary software toolkits such as TensorRT provide ways of customization, but they are often insufficient to satisfy software developers’ needs. AITemplate can help with that.

Deloitte Acquires SFL Scientific to Enhance AI

By

Disha Chopra

-

October 4, 2022

Deloitte, an industry leader in consulting and financial services provision, acquires SFL Scientific, an award-winning AI strategy firm, to enhance AI. The acquisition is a part of Deloitte’s dedication to enhancing consumer experience by technologically transforming the company.

Integrating SFL Scientific’s expertise in AI and analytics with Deloitte’s capabilities, industry depth, and experience strengthens the latter’s position, improves its capacity to carry out AI-driven transformations for clients, and hastens the growth of its US AI market share.

Dan Helfrich, CEO and chairman of Deloitte Consulting LLP, said, “As our clients advance their enterprise AI journeys, they require the comprehensive and sophisticated experience that SFL Scientific delivers.”

SFL Scientific has a record of effectively executing high-quality, unique, and challenging projects while providing scientific capabilities to the private and governmental sectors. The skilled AI strategists at SFL Scientific serve as trusted consultants to CEOs, guiding them to comprehend and assess emerging areas for AI investment and spot opportunities to improve the companies.

Helfrich added that the acquisition would reinforce the company as a career destination leveraging artificial intelligence to foster a technically growing environment of data scientists, AI strategists, and technologists.

Hoopr.ai Launches the #HarGharCreator Campaign to Empower Content Creators

By

Disha Chopra

-

October 4, 2022

Hoopr.ai, a music licensing platform, launches the #HarGharCreator campaign to empower content creators and help them find the right Indian music for their videos. In a commercial, a teenage boy tells his conservative father he would like to be a content creator rather than an engineer. But the father does not seem to budge, given the prejudice against the former. Hoopr.ai wants to redefine how people view creators with their #HarGharCreator campaign.

In December 2021, Hoopr.ai received $1.5 million seed funding from backers, including Venture Catalysts and Ashneer Grover. The company offers music tracks by musicians like Monali Thakur, Ash King, Nikhil D’souza, and others in Hindi, English, Punjabi, and other regional languages. After going live, Hoopr.ai reports that more than 15,000 creators use the platform.

According to the company, India has a creator community of over 80 million people who produce audio-visual content on various subjects. This community includes vloggers, podcasters, gamers, filmmakers, live-streamers, and influencers.

According to the company, the number of people who directly depend on the creator ecosystem for their livelihood is currently over six lakh, which is expected to rise. Additionally, people are more aware of concerns like copyrights. Thus, there is an increased awareness of the necessity for licensed music.

Gaurav Dagaonkar, CEO and co-founder of Hoopr.ai, said, “Apart from individual creators, the music on Hoopr.ai is also being used by brands, enterprises, and OTT platforms.” He also added that Hoopr’s music is wholly licensed and free from copyright strikes or potential takedowns.

Meta announces date for Facebook Communities Summit 2022

By

Sahil Pawar

-

October 4, 2022

Meta announces date for Facebook Communities Summit 2022

Meta has announced the date for its Facebook Communities Summit 2022, with a virtual event scheduled for October 20th. Meta will bring together moderators from various Facebook groups to provide critical insights and tips while sharing previews of upcoming groups’ tools and features.

Previous Communities Summit events have seen the launch of various new tools and updates, so it could be worth tuning in if one is looking to stay up with the latest group moderation options that could help maximize one’s Facebook community.

Facebook Groups remain a vital element of the broader Facebook experience. Despite the app losing engagement, groups are a key interaction space for many people on various niche topics. With the main feed becoming so divisive and political, many users have switched to groups to avoid misinformation and engage with like-minded people in a more private, enclosed way.

Meta has repeatedly noted that Facebook users are not interested in the politically divisive content in the app, and groups offer some respite from this. And with almost 2 billion people still logging onto Facebook every day, it can be an important connective consideration and a good alternative for building community.

Meta has also been working on adding more monetization options for groups, which could provide new considerations in meeting one’s target audience where they are spending their time. As such, it could be worth tuning into the Communities Summit later this month to get the latest info on Facebook Groups and where things are headed.

Using AI to Target, Shoot & Kill Cockroaches

By

Disha Chopra

-

October 4, 2022

Researchers from Heriot-Watt University, University of Sussex, and University Paul Sabatier have developed a technology using AI to target, shoot, & kill cockroaches. IIdar Rakhmatulin, Mathieu Lihoreau, and Jose Pueyo have described the tool in Oriental Insects, a journal based on insects.

With varying degrees of success, numerous attempts have been made to develop products intended to kill roaches. Insecticides can be dangerous for people, pets, and the environment, which is a significant disadvantage to most such products. The researchers’ brand-new solution to the issue in this new endeavor is to kill people with a laser beam.

The device is a straightforward machine learning-driven device. It was based on Jetson Nano, a compact device with machine learning software, to which the researchers added a galvanometer, a programmable laser, and two cameras. The galvanometer was used to receive data from the AI unit and then utilize that data to adjust the laser’s direction.

They discovered that the equipment could precisely detect and kill cockroaches on testing the prototype. They also discovered that, like the “Star Trek” phaser, they could fine-tune the laser to allow different impacts.

If they wanted, the researchers could stun the cockroach, which usually causes the cockroach to change its direction. They could also set the laser to kill, in which case it would execute its intended target.

The photos used for training have been uploaded on GitHub, and the tracking dataset has been posted on Kaggle. The researchers maintain that they have no intent to market the device. Anyone can use the method described in their study to create a device of their own.

ONDC opens beta testing to consumers in 16 locations across Bengaluru

By

Sahil Pawar

-

October 3, 2022

ONDC opens beta testing to consumers in 16 locations across Bengaluru

According to the Ministry of Commerce and Industry, Open Network for Digital Commerce (ONDC), an initiative of the Department of Promotion of Industry and Internal Trade (DPIIT), has opened its network to consumers in 16 locations across Bengaluru.

To start with, consumers can place their orders in two domains – groceries and restaurants through buyer apps participating in the ONDC network. Currently, Spicemoney, Mystore, and PayTM are available as buyer apps. Consumers in the city can buy grocery products or order food from the stores and restaurants through seller apps Innobits Mystore, nStore, SellerApp, Ushop, e-Samudaay, eVitalrx, Bizom, Digiit, Go Frugal, Growth Falcons, and Uengage.

ONDC’s beta test in Bengaluru is a significant first step in operationalizing a network approach to e-commerce, an alternative to a platform-centric approach. The ministry said this would make the e-Commerce landscape more inclusive, accessible, and experience-driven for all consumers and sellers.

Consumers can now shop for various categories of services and products from a single buyer application. They can order food or buy grocery products from stores and restaurants. In the coming weeks, more applications will join the network, expanding sellers and buyers who can participate in this transaction.

ONDC began its Alpha test phase with a selected user group of buyers first in Bengaluru in April, expanding to over 80 cities by September 2022. This validated the apps and confirmed operational flows and business. Now in the beta testing phase, the public will be able to experience shopping through ONDC and give early feedback for necessary action before expanding further.

To build trust amongst sellers, consumers, and Network Participants in the open network, ONDC has been consulting current ecosystem experts and examined best practices for adapting and evolving the best approach to build trust in an unbundled decentralized network. This is explained in a consultation paper being made available for public consultation on the official social media handles of ONDC, DPIIT, and the website.

Musk says 160,000 Tesla customers are now running the FSD beta

By

Sahil Pawar

-

October 3, 2022

160,000 Tesla customers are now running the FSD beta

Elon Musk announced that 160,000 Tesla customers are now running the Full Self Driving (FSD) beta, from a mere 2,000 last year. At Tesla AI Day 2022 on September 30, Musk said the FSD technology would be ready for a global roll-out by December this year.

So far, Tesla has released 35 software updates of FSD, which have generated regulatory scrutiny after several crashes were reported. Musk acknowledged the regulatory and testing hurdles that need to be sorted before a global roll-out is planned.

Currently, FSD is available to around 160,000 drivers in Canada and the US. Recently, a Tesla owner sued the electric car maker, saying the company and its CEO Musk are misleadingly marketing the Autopilot and Full Self-Driving software.

The owner alleged that Tesla has been deceiving and misleading consumers regarding the current capabilities of its ADAS technology by portraying that it was constantly on the cusp of perfecting that technology to produce a fully self-driving car.

Tesla has come under heavy scrutiny over the controversial Autopilot advanced driver assistance system that has allegedly killed many, with US state and federal regulators turning up the heat on the company.

The Department of Motor Vehicles (DMV), California, also accused Tesla of running fake claims about its Autopilot and FSD features last month.

Tesla unveils latest version of Dojo supercomputer at Tesla AI Day 2022

By

Sahil Pawar

-

October 3, 2022

Tesla unveils latest version of Dojo supercomputer

Tesla has unveiled the latest version of the Dojo supercomputer, and it is so powerful that it tripped the power grid in Palo Alto. Dojo is Tesla’s custom supercomputer platform built from the ground up for AI machine learning and, more specifically, video training using the video data from its fleet of vehicles.

The automaker already has a giant NVIDIA GPU-based supercomputer that is one of the most powerful in the world. However, the new Dojo custom-built computer uses chips and an entire infrastructure designed by Tesla.

The custom-built supercomputer will elevate Tesla’s capacity to train neural nets using video data, which is critical to its computer vision technology powering its self-driving effort.

Last year, at Tesla’s AI Day, the company unveiled its Dojo supercomputer, but it was still ramping up its effort at the time. It only had its first chip and training tiles, and it was still working on building a full Dojo cabinet and cluster, or “Exapod.”

Tesla has unveiled the progress made with the Dojo program over the last year during its AI Day 2022 on September 30. The company confirmed that it managed to go from a chip and tile to a system tray and a full cabinet.

Tesla claimed it could replace six GPU boxes with a single Dojo tile, which the company claims costs less than one GPU box. There are six of those tiles per tray. Tesla said that a single tray is the equivalent of 3 to 4 fully-loaded supercomputer racks. The company is integrating its host interface directly on the system tray to create a sizeable full host assembly. Tesla can fit two of these system trays with host assembly into a single Dojo cabinet.