So this is what the tech world has been up to this week, interesting happenings all over. We have the latest in tech for this week, week 35 of 2023:
#1. Open AI launches ChatGPT Enterprise
OpenAI launched ChatGPT Enterprise for companies seeking to leverage the wildly successful AI-powered chatbot for business purposes. The Enterprise version is based on GPT-4 and can still perform tasks like debugging code and composing emails, but with extra enterprise-level features. These include an admin console for employee management, sharable prompt templates for creating workflows, and unmatched customizations to create AI-powered tools. It also grants businesses unlimited access to ChatGPT’s Advanced Data Analysis feature. It allows users to analyze data from uploaded files. These features match the recently launched Bing Chat Enterprise offering by Microsoft.
While OpenAI does not disclose pricing plans for ChatGPT Enterprise, the company says that the solution is intended to complement ChatGPT Plus, not replace it. At this time, users can pay $20 per month for ChatGPT Plus and access GPT-4 and Advanced Data Analysis. Enterprise offers priority access to GPT-4, meaning that companies get faster performance and expanded context windows to support 25,000 words. OpenAI also emphasized that business data will remain encrypted, so ChatGPT will not use internal information to train models.
ChatGPT Enterprise is OpenAI’s attempt at improving monetization. Official reports indicate that the company spent over $540 million to develop the chat bot, and it costs $700,000 per day to run. The company also earned $30 million in revenue in 2022, which shows a glaring imbalance in its financial outlook. ChatGPT Enterprise promises more powerful capabilities for organizations, but that remains to be seen as the free and Plus editions continue to deliver efficient solutions for users.
#2. X acquires approval for Rhode Island crypto license
The social media platform X (formerly Twitter), received the go-ahead to introduce cryptocurrency solutions into its platform. The license granted by the Rhode Island Currency Transmission authorizes X to store, transfer, and trade digital currencies. It turns X into a digital wallet, enabling X’s 400 million users to potentially transact on the platform via crypto. This is a leap forward for X as its current owner, Elon Musk, has expressed a desire to convert the platform into an all-inclusive app. With crypto integration, X users could transfer funds, tip content creators, make in-game microtransactions, and more on the platform.
Uncertainties still surround this development. For example, regulatory standards surrounding crypto may restrict users from certain countries from engaging with X in this way. There is also speculation about Dogecoin becoming X’s chosen cryptocurrency because of Musk’s involvement in it. Investors can only wait for Musk’s updates concerning this revolutionary move.
#3. DeepMind tests SynthID
Google’s AI research laboratory, DeepMind, is on track to identify AI images using a digital watermark, SynthID. The watermark alters individual pixels in machine-generated images such that they’re detectable by computers but not to the naked eye. SynthID is Google’s response to disinformation because it becomes increasingly difficult to identify real images from AI-generated ones.
Tools like Midjourney enable millions of users to create images within seconds using text-based prompts. Users can easily crop or edit out watermarks from these AI-generated images, leading to ethical and legal concerns about ownership and copyright. With SynthID, users can submit images they suspect to be AI-generated and find out instantly if they’re real or machine-made.
However, the launch of SynthID comes with caveats. First, it will only apply watermarks to AI images created with Google’s text-to-image tool, Imagen AI. DeepMind also clarifies that SynthID is still in the experimental phase, so it’s not completely foolproof. But it is a solid step toward fulfilling Google’s commitment to the safe use of AI.
Other major tech companies like Amazon, Microsoft, and Meta have also pledged to use some form of watermarking technology for AI-generated content. These companies are far from reaching a standardized AI-monitoring method, further complicating the process of achieving transparency in AI content.
#4. Google’s Pixel 8 Pro design leak
An image of the Pixel 8 Pro phone appeared on the Google Store website this week, but it was quickly taken down. The leak indicates that the much-anticipated successor to the Pixel 7 Pro has a triple rear camera and a Porcelain White housing. The Pixel 8 Pro shows a slightly different camera housing and another sensor near the LED flash. The image, which was discovered and shared by an X (formerly Twitter) user, was too low-resolution to definitively identify these and more features.
The highly-anticipated Pixel 8 handsets are set to officially launch on October 3, 2023. They are expected to be eSIM-only devices to rival Apple’s iPhone 14 and 14 Pro models. The Pixel 8 and 8 Pro will also receive five years of OS updates, overtaking Oppo, Samsung, and OnePlus devices in software support.
#5. US-EU Partnership seizes global malware network
FBI, Europol, and other law enforcement partners seized more than 50 servers hosting the malware loader, Qakbot, in an operation dubbed “Duck Hunt.” During the takedown, the authorities removed the malware agent remotely to disinfect thousands of computers and dismantled the network. They also seized $8.6 million in cyber currency but made no arrests, since the Qakbot administrators are believed to be in Russia or other former Soviet countries. The investigation is still ongoing as cybersecurity experts warn that this setback to the hackers would only be temporary.
Qakbot, also known as Qbot and Pinkslipbot, first appeared in 2008 and has been used to commit devastating ransomware attacks around the world. In 2023 alone, Qakbot was the most commonly detected malware impacting one in 10 organizations, resulting in 30% of global cyber attacks. Qakbot earned an estimated $58 million from ransomware attacks within the last 18 months. The FBI called the network one of the “most devastating cybercrime tools in history.”
Qakbot typically uses phishing emails to grant initial access into a network, then deploys malicious payloads to steal user information and facilitate a range of fraudulent and criminal activities. Qakbot has disrupted local governments, businesses, hospitals, and schools worldwide.
This Qakbot takedown is an impressive feat for the FBI and its global partners, following the Hive ransomware operation in January this year.
