Dark web ChatGPT, robots and Railtown

Railtown — An AI copilot for software developers

Welcome back to The Daily Update — Here’s what’s on deck today:

🗞️ The Latest in AI: FraudGPT uses generative AI for cybercrime

🤖 AI Tool of the Day: Railtown — An AI copilot for software developers

📸 Business Spotlight: Google DeepMind shows off new system for teaching robots novel tasks

🚨 AI Roundup: Four quick hits

Read time: 3.5 minutes

🗞️ The Latest in AI: FraudGPT uses generative AI for cybercrime

Source: The Hacker News

A new cybercrime generative AI tool called FraudGPT is being advertised on the dark web and Telegram channels. The tool has already gained over 3,000 confirmed sales and reviews.

The chatbot is designed for offensive purposes, such as crafting spear phishing emails and creating undetectable malware. The all-in-one cybercrime tool can write malicious code, create phishing pages and find leaks and vulnerabilities.

FraudGPT’s original publisher goes by the online alias CanadianKingpin. The tool has been offered since at least July 22 for a subscription cost of $200 a month.

“If your [sic] looking for a ChatGPT alternative designed to provide a wide range of exclusive tools, features, and capabilities tailored to anyone’s individual needs with no boundaries then look no further!,” CanadianKingpin said.

Security experts are emphasizing the need for continuous innovation in cybersecurity to combat threats like FraudGPT. The emergence of FraudGPT is likely just the beginning of the spread of cybercrime generative AI tools.

Click here to learn more.

🤖 AI Tool of the Day: Railtown — An AI copilot for software developers

Railtown is an all-in-one AI copilot for developers. It leverages AI and machine learning to monitor, diagnose and prioritize errors efficiently. Railtown features three AI copilot solutions:

Root Cause Analysis (RCA)

This copilot notifies you of errors in your CI/CD environment in real time. It identifies any code change that is the root cause of an error, issue, bug or exception. Railtown’s AI analyzes how each error affects your application and prioritizes them so you know what to act on first. This tool saves developers over eight hours per week.

Release Notes

Your team can now create comprehensive release notes with one click. Simply connect your ticketing platform and Railtown’s AI technology will automatically generate release notes. The Release Note copilot creates a concise executive summary in a descriptive business format with the intent of consumption by end users, engineers, team leads, marketing and management. You can generate your first draft in under two minutes.

Scrum Master (coming soon)

This knowledge-based chat copilot is trained on a targeted language model (TLM) built on any software application connected to Railtown. This allows the chat copilot to provide developers with accurate answers to any questions about an application. Developers can ask the copilot specific questions or describe the problem they are facing, and it will provide relevant and actionable information to help solve the problem.

Click here to learn more.

📸 Business Spotlight: Google DeepMind shows off new system for teaching robots novel tasks

Source: Google DeepMind

Google DeepMind just unveiled its newly launched vision-language-action model Robotic Transformer 2 (RT-2). The model allows robots to understand human instructions and translate them into actions.

RT-2 is trained on text and images from the web, meaning it can perform different tasks without explicit training.

For example, in a scenario in which a robot is asked to throw away trash, many other models require a user to teach the robot what qualifies as trash and then train it to pick up garbage and throw it away. RT-2 is able to recognize trash and dispose of it on its own.

Google DeepMind’s robotics team says the ability to execute new tasks has improved from 32% to 62% from RT-1 (previous model) to RT-2.

Click here to learn more.

🚨 AI Roundup: Four quick hits

  • Dell is all in on generative AI.

  • FBI: AI is making it easier for hackers to write malware.

  • MIT develops PhotoGuard to protect images from AI manipulation.

  • Adam Selipsky: There will not be one generative AI model that rules them all.

Have questions or thoughts on today’s newsletter? Reply to this email and I’ll get back to you as soon as I can. Thanks for reading The Daily Update!

Jack

(P.S. If you know someone that may find today’s newsletter useful, send it their way!)