DoorDash is leveraging its extensive network of 8 million U.S. couriers to train artificial intelligence and robotics through a new app called Tasks. Couriers can earn money by recording videos and completing digital tasks, such as filming household chores or conversations. This initiative aims to gather crucial real-world data that helps AI models understand physical tasks, a capability simulations often cannot replicate. This strategy mirrors efforts by other gig economy companies like Uber and Instacart, with DoorDash planning to expand task types and locations.
In the realm of enterprise AI, Salesforce has launched Agentforce Sales, an AI-powered platform designed to automate various sales tasks. This platform integrates autonomous agents that can manage lead qualification, pipeline, and proposal generation around the clock. Salesforce aims to compete with offerings from companies like Microsoft, ServiceNow, and SAP by deeply embedding this solution into its CRM system. Meanwhile, Nokia is looking ahead to Wi-Fi 9, envisioning it as a standard focused on predictable, low-latency performance essential for AI-driven systems like robotics and extended reality, rather than just faster speeds.
The application of AI is also expanding into creative and specialized fields. Pete Hottelet, founder of Omni Consumer Products, utilized Anthropic's Claude AI to build a 3D interactive replica of the Voight-Kampff machine from Blade Runner 2049 in about 90 minutes, demonstrating AI's ability to handle procedural tasks without manual coding. This highlights the growing importance of human skills like creativity and communication. Additionally, Amjad Masad, CEO of Replit, suggests that not knowing how to code could be an advantage in the AI era, as AI tools lower the barrier to software creation, making product vision and market understanding more critical.
Further developments include Squad, a new tool that coordinates specialized AI agents within a code repository, allowing teams of AI to collaboratively implement features, write tests, and review work. On a more personal level, AI has been used to design a cancer treatment for a dog by an Australian tech entrepreneur, showcasing its potential in personalized veterinary medicine. Food tracking apps, often AI-powered, also offer increased awareness and accountability for eating habits, though users should be mindful of potential perfectionism.
Discussions around AI's broader impact continue, with a recent Hard Fork episode exploring tech layoffs at companies like Atlassian and Block, questioning if AI is the primary cause. The episode also touched on why chatbots struggle with creative writing and the trend of 'tokenmaxxing,' where companies measure AI usage. These conversations underscore the evolving understanding of AI's capabilities, limitations, and its influence on the workforce and creative industries.
Key Takeaways
- DoorDash is using its 8 million U.S. couriers and a new Tasks app to collect audio and video data for AI and robotics training, paying couriers for tasks like household chores.
- Salesforce launched Agentforce Sales, an AI-powered platform with autonomous agents to automate sales tasks, competing with Microsoft, ServiceNow, and SAP.
- Nokia is advocating for Wi-Fi 9 to prioritize predictable, low-latency performance (sub-10 ms) for AI-driven systems like robotics and extended reality.
- Pete Hottelet of Omni Consumer Products used Anthropic's Claude AI to create a 3D interactive Blade Runner machine replica in 90 minutes without manual coding.
- Amjad Masad, CEO of Replit, believes not knowing how to code can be an advantage in the AI era, emphasizing product vision over technical coding skills.
- Squad is a new tool that coordinates specialized AI agents (lead, frontend, backend, tester) to work collaboratively within a code repository for software development.
- An Australian tech entrepreneur utilized AI to design a cancer treatment for his dog, highlighting AI's potential in personalized veterinary medicine.
- AI-powered food tracking apps help users monitor calorie intake and nutrients, though accuracy can vary, and professional guidance is recommended.
- A Hard Fork episode discussed tech layoffs, questioning AI's role, and explored why chatbots struggle with creative writing and the 'tokenmaxxing' trend.
- The use of gig workers for AI data collection is a growing trend, with DoorDash's initiative focusing on real-world physical task data.
DoorDash couriers now train AI with paid tasks
DoorDash is using its millions of delivery couriers to help train artificial intelligence and robotics. Through a new app called Tasks, couriers can earn money by recording videos and completing other digital tasks. This data will help improve AI models for DoorDash and its partners in various industries. Similar efforts have been made by companies like Uber and Instacart. The tasks include filming household chores, recording conversations, and taking pictures for digital menus. DoorDash is offering these paid opportunities in select U.S. markets, with plans to expand to more task types and countries.
Gig workers become AI training data providers
The gig economy is now a key source for training artificial intelligence. DoorDash has launched a new app called Tasks, using its 2 million U.S. couriers to collect audio and video data for AI and robotics training. This data helps train AI for real-world physical tasks that simulations cannot replicate. Competitors like Uber are also using their workforces for data collection. Tasks include recording everyday chores and conversations. DoorDash plans to expand these data collection efforts to more task types and locations.
DoorDash drivers earn money training AI robots
DoorDash is now allowing its drivers to train AI robots as a new side gig. The company launched a standalone app called Tasks, where its 8 million U.S. gig workers can record themselves performing various tasks for payment. This data helps AI and robotics models understand the physical world better. Tasks include household chores like folding clothes and washing dishes, as well as recording conversations in different languages. DoorDash cofounder Andy Fang believes this will significantly advance physical intelligence. The company plans to add more task types over time.
DoorDash uses 8 million couriers for AI training
DoorDash is transforming its 8 million couriers into an AI training resource through its new Tasks app. This app allows couriers to earn money by submitting raw audio and video footage of tasks like household chores. This data is crucial for training AI and robotics models, especially for physical tasks. The program is currently available in select U.S. markets, excluding areas with stricter regulations. This move follows a similar strategy by Uber, but DoorDash focuses more on physical-world video data. The demand for human-annotated data is rapidly growing, making this a valuable revenue stream for DoorDash.
Nokia envisions Wi-Fi 9 for AI and real-time needs
Nokia is shaping the future of Wi-Fi by focusing on real-time connectivity for AI-driven systems, not just faster speeds. The company is advocating for Wi-Fi 9 to prioritize predictable, low-latency performance essential for applications like robotics and extended reality. Nokia aims to influence the IEEE standards to ensure Wi-Fi 9 supports deterministic performance, with sub-10 ms latency and consistent multi-gigabit delivery. This vision aligns Wi-Fi 9 with 6G to create a synchronized wireless ecosystem. The focus is on energy efficiency and reliable performance in dense environments.
AI helps create cancer treatment for dog
An Australian tech entrepreneur used artificial intelligence to design a cancer treatment for his dog. Paul Conyngham, a data analyst, shared how he utilized AI to develop the treatment. The story highlights the potential of AI in veterinary medicine and personalized treatments. This innovative approach showcases how AI can be applied to solve complex health challenges, even for pets.
AI layoffs, chatbot writing, and tokenmaxxing discussed
This episode of Hard Fork discusses the latest wave of tech layoffs at companies like Atlassian and Block, questioning if AI is truly the cause. It also features writer Jasmine Sun discussing why chatbots struggle with creative writing. The hosts explore 'tokenmaxxing,' a trend where tech companies create leaderboards to measure AI usage. The conversation touches on the impact of AI on jobs and the current limitations of AI in creative fields. The episode aims to provide insights into these evolving AI-related topics.
Salesforce launches AI sales platform Agentforce Sales
Salesforce has launched Agentforce Sales, an AI-powered platform designed to automate and improve sales tasks. This new platform embeds autonomous agents that can handle lead qualification, pipeline management, and proposal generation 24/7. It aims to compete with offerings from Microsoft, ServiceNow, and SAP by integrating deeply into Salesforce's CRM. The platform is available as an add-on with tiered pricing. The success of Agentforce Sales may depend on enterprise adoption of integrated AI solutions versus modular tools, considering potential concerns about cost and trust.
AI builds interactive Blade Runner machine, highlights human skills
Omni Consumer Products founder Pete Hottelet used AI to build a 3D interactive replica of the Voight-Kampff machine from Blade Runner 2049. The project was completed in about 90 minutes using conversational prompts with Anthropic's Claude AI, without manual coding. This demonstrates how AI can handle procedural tasks, shifting the focus to essential human skills like creativity, communication, and tenacity for the future workforce. As AI automates more tasks, these human qualities become critical for hiring and innovation.
Squad coordinates AI agents within your code repository
Squad is a new tool that allows coordinated AI agents to work directly within your code repository. Instead of a single chatbot, Squad deploys a specialized team of AI agents, including a lead, frontend developer, backend developer, and tester. These agents access repository context and shared decisions to work collaboratively. For example, one agent can implement features while another writes tests, with a separate agent reviewing the work. This approach ensures genuine independent review and improves the efficiency of software development by managing context and iteration internally.
Replit CEO: Not knowing how to code is an advantage
Amjad Masad, CEO of the AI-powered coding platform Replit, believes that not knowing how to code can be an advantage in the current AI era. He suggests that individuals focused on product, marketing, and user interface, rather than coding details, are better positioned to succeed. Masad predicts that by this year, entrepreneurs without coding backgrounds will have an advantage. This shift is driven by AI tools that lower the barrier to software creation, making idea generation and market understanding more critical than technical coding skills.
Food tracking apps offer awareness and accountability
Food tracking apps, often powered by AI, can significantly increase awareness and accountability around eating habits. These apps help users monitor calorie intake, nutrient consumption, and meal timing. Some apps use AI camera analysis to estimate meal calories, though accuracy can vary. While helpful for goals like weight loss, users should be mindful of potential obsession or anxiety related to perfectionism. Consulting a registered dietitian is recommended for personalized calorie and nutrition guidance beyond app-based calculations.
Sources
- DoorDash taps millions of couriers to train artificial intelligence
- The Gig Economy Is Now the Training Layer for AI
- DoorDash is now letting its drivers train AI on the side
- DoorDash Is Turning 8 Million Couriers Into An AI Training Machine
- Nokia Sets Early Wi-Fi 9 Vision Around AI
- Australian tech entrepreneur uses AI to create cancer treatment for his dog
- ‘A.I.-Washing’ Layoffs? + Why L.L.M.s Can’t Write Well + Tokenmaxxing
- Can Agentforce Sales Redefine AI Sales, Or Will Platform Fatigue Slow Adoption?
- Omni Consumer Products Founder Builds 3D, Fully Interactive, Blade Runner Voight-Kampff Machine in 90 Minutes Using AI, Highlights Three Human Skills That Will Define the Future Human Workforce
- How Squad runs coordinated AI agents inside your repository
- Knowing How To Code Can Be A Disadvantage In The AI Era: Replit CEO Amjad Masad
- I Learned More Than I Thought I Would From Using Food-Tracking Apps
Comments
Please log in to post a comment.