What is one specific way generative AI has changed how work gets done in your field, and why?
Most answers highlight generative AI accelerating work, automating tasks, and shifting roles toward curation, prompting, or oversight, though one answer is skeptical of AI's value.
19 answers
Answers


Benedict Evans
Independent analyst
Smaller teams, faster projects
Benedict Evans says generative AI lets work be done by smaller teams on faster projects, citing Mark Zuckerberg's example of tasks that once took a dozen people a couple of months now taking a small number of people a couple of weeks.
Antler Global · Watch on YouTube
Dylan Babbs
Co-founder, CTO · Profound
100× Faster Building
Dylan Babbs says generative AI has made designers 100 times more efficient and faster at building and designing, while the core skills of understanding the business, visual design and building remain unchanged.
First of Kind · Watch on YouTube

Shayde Christian
Chief Data & Analytics Officer · Cloudera
Natural Language Querying
Shayde Christian says natural language querying automates his data analytics and data analyst role, letting the business do it, and argues analysts remain valuable because they know the data better and are better with AI.
The answer starts at 41:57.
Chain of Thought | AI Agents, Infrastructure & Engineering · Listen on Apple Podcasts

Carrol Chang
Chief Executive Officer · Andela
AI coding agents write code
Carrol Chang says some software engineers now give AI directions, context and guardrails while coding agents write the code, and the engineer reviews the output, because AI has enabled that.
The answer starts at 3:46.
The AI in Business Podcast · Listen on Apple Podcasts

Marty Cagan
Partner · Silicon Valley Product Group
Constant shared document
Marty Cagan says AI makes it easy to keep a constant shared document of everything known about the product, giving the whole team access, which he says makes much of the past communication overhead redundant.
Mind the Product · Watch on YouTube

Kieran Klaassen
GM of Cora | EIR · Every Inc.
Teaching AI personal context
Kieran Klaassen says people overcomplicate AI writing with prompt libraries and brand voice documents, and argues the real advantage comes from teaching AI about you through small context files rather than more about writing.
Kieran’s Substack - The AI Marketing Generalist · Kieran’s Substack - The AI Marketing Generalist · Read

Sam Altman
Chief Executive Officer · OpenAI
Persistent AI agents
Sam Altman describes a move toward persistent intelligence, with agents running in the background 24/7 connected to your systems and doing things proactively, which he says feels important to explore.
Every · Watch on YouTube

Eric Schmidt
Chair and CEO · Relativity Space
AI code generation
Eric Schmidt says AI code generation has changed his field by letting him act as an architect who describes what he wants while the code gets written, rather than building it himself.
Blackstone · Watch on YouTube

Molly Graham
Host, WorkLife Podcast · Ted Conferences
Talking to AI to write code
Molly Graham says engineering work has transformed: writing code in an IDE is no longer part of the job, and now it's talking to AI to write the code, whether you're there or not.
Lenny's Podcast · Watch on YouTube

Anastasis Germanidis
Co-Founder & CTO · Runway
Real‑time iteration
Anastasis Germanidis says people will increasingly care about latency as models improve, since the ability to iterate very quickly becomes more important, letting users explore faster by generating many outputs almost instantly.
Latent Space · Latent Space · Read

Katie Dill
Head of Design · Stripe
Using specific prompts
Katie Dill recommends putting in specific prompts, because improving your inputs with your brand, interests and source material makes AI a stronger creative partner.
Lenny's Podcast · Watch on YouTube

Nikhil Mungel
Head of AI Research & Development · Cribl
Accelerated software creation speed
Nikhil Mungel says agentic AI coding has accelerated how everyone creates and produces software, which means there has been a lot more software.
The answer starts at 3:50.
DataFramed · Listen on Apple Podcasts
Jonathan Evens
Product Lead · Google Deepmind
Prompt‑based development replaces hyperparameter tuning
Jonathan Evens says the field has moved from traditional machine learning engineering, with hyperparameter tuning, to prompt-based work, where much can be done with prompting and iterating on prompts and eval sets.
Mind the Product · Watch on YouTube

Reid Hoffman
Co-Founder, Executive Board Chair · Manas Ai
AI generates many options, human curates
Reid Hoffman says generative AI can produce many options, such as 200 images or names, and that the human's role is curation—picking the one that fits, which is where their taste and creative wisdom come in.
Reid Hoffman · Watch on YouTube

Amjad Masad
CEO · Replit
Using AI agents to augment skill gaps
Amjad Masad says that wherever he has a skills gap, he can use agents to augment it, describing this as part of how he benefits from a broad set of skills.
N/A · N/A · Listen at 6:38

Ed Zitron
Founder/CEO
AI Is A Con
Ed Zitron does not give this answer in the passage; the plain-English question line is sponsor copy about NetSuite, and his own remarks call generative AI a con.
The Diary Of A CEO with Steven Bartlett · The Diary Of A CEO with Steven Bartlett · Listen

Ian Silber
Head of Product Design · Openai
AI agents for instant prototyping
Ian Silber says people embracing AI use it at every step, suggesting they throw an idea into a tool like Codex or Claude and into an agent to help think about and prototype it, as a different way of working.
Lenny's Podcast · Watch on YouTube
Dipankar Mazumdar
Director - Data/AI Developers · Cloudera
Code generation
Dipankar Mazumdar says generative AI is making data engineers more productive, particularly around code generation, documentation, debugging and understanding unfamiliar systems, because it lowers the barrier to getting started and helps engineers move faster on repetitive tasks.
Junaid Effendi | Sharing knowledge for Engineers · Junaid Effendi | Sharing knowledge for Engineers · Read

Kashyap Dalal
Chief Operating Officer · Simplilearn
Generating content on demand
Kashyap Dalal says AI's ability to generate content on demand means learning content must become dynamic, since courses go obsolete quickly and can no longer be sold unchanged for two years.
Analytics Insight · Analytics Insight · Listen at 12:17