/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Some AI companies are hiring “prompt engineers”, who create and refine text prompts for chatbots to understand the AI systems' flaws and coax optimal results

When Riley Goodside starts talking with the artificial-intelligence system GPT-3, he likes to first establish his dominance.

Washington Post Drew Harwell

Context & Ripple Effects

Two years into the ChatGPT era, AI companies have turned coaxing models into a job title: the prompt engineer, exemplified by Riley Goodside, who probes GPT-3 for its flaws and refines text inputs until the system yields its best output. The role emerged alongside a broader scramble in which executives, engineers, and other employees began experimenting with ChatGPT to speed up tasks or avoid being left behind.

What makes this hiring wave notable is how explicitly temporary it looks from inside the same coverage arc: by 2025, prompt engineering is reported obsolete as models get better at intuiting user intent and companies fold the skill into ordinary staff training. The specialist job and its dissolution are two ends of the same story.

First-order effects

  • Companies building on GPT-3-class models gain a dedicated role whose output is quality control on the interface itself — finding flaws and extracting optimal results that casual users can't reach.

Second-order effects

  • The prompt layer becomes something actors outside the lab actively manipulate: the related reporting on a reporter adding a "strategic text sequence" to data sources shows prompts turning into an optimization battleground, not just a craft.

Third-order effects

  • If the 2025 obsolescence pattern holds, the role gets absorbed rather than eliminated — models intuit intent directly while employers train general staff on prompting, echoing later findings that even "AI native" hires need careful oversight of their tool use.

The trend: Prompt engineering's rise and rapid obsolescence mark the model layer absorbing the prompt layer, as interfaces stop requiring specialists to translate human intent.

Discussion

  • @drewharwell@mastodon.social Drew Harwell on mastodon
    New: Tech's hottest new job: AI whisperer.  No coding required.  Prompt engineers sweet-talk GPT-3 and other generative-AI tools using plain English - a revolution in how humans work with machines. …
  • @jeffjarvis Jeff Jarvis on x
    What is a computer programmer but a prompt engineer? What is a commissioning editor but a prompt engineer? What is a a teacher's assignments but prompt engineering? Yes, it's a good skill to learn: saying exactly what you want. https://www.washingtonpost.com/ ...
  • @neilturkewitz Neil Turkewitz on x
    “The AI start-up Anthropic, founded by former OpenAI employees...recently listed a job opening for a “prompt engineer & librarian” with a salary ranging up to $335K.” All on the back of conscripted labor of artists. It's time to cry. ⁦@drewharwell⁩ https://www.washingtonpost.com/…
  • @willoremus Will Oremus on x
    Tech's hot new job: “prompt engineer.” Chatbot wranglers like @goodside claim expertise in how to bend ChatGPT, Bing, Dall-E et al. to their will—no coding required. Fascinating story by @drewharwell (whose Twitter account is still locked by Musk): https://www.washingtonpost.com/…
  • @nitashatiku @nitashatiku on x
    new @drewharwell dispatch from the “early weirdness” of LLMs: profile of prompt engineer @goodside and getting paid to talk to machines “It's a very good tool, he tells it, but it's not perfect, and it needs to obey whatever he says.” https://www.washingtonpost.com/ ...
  • @lukew Luke Wroblewski on x
    sounds like the creative process to me... “he spent 80 hours perfecting [his prompt] over 900 iterations.” source: https://www.washingtonpost.com/ ...
  • @manan @manan on x
    “engineers” is an interesting choice here. https://www.techmeme.com/...
  • @alexandr_wang Alexandr Wang on x
    it's been a pleasure working with @goodside. as time has gone on, it feels more and more certain that AI whispering is here to stay. https://www.washingtonpost.com/ ...
  • @anthony Anthony DeRosa on x
    The ability to ask AI just the right questions is a highly sought after, well paid skill: https://t.co/...
  • @taylorlorenz Taylor Lorenz on x
    ‘Prompt engineers’ are being hired for their skill in getting AI systems to produce exactly what they want. And they make pretty good money. https://www.washingtonpost.com/ ...
  • @dzorich Diane Zorich on x
    “The best creations (for AI image creators) rely in humans specialized knowledge from fields such as art history & graphic design.” New alt-ac career path for art historians? (move these tools away from the largely kitsch products they currently output) https://www.washingtonpost…
  • @farzad_md Farzad Mostashari on x
    A small set of obscure terms, such as “SolidGoldMagikarp” tended to induce what they called a “mysterious failure mode” most notably, a garbled stream of profane insults. They're still not entirely sure why Tech's hottest new job-AI whisperer @drewharwell https://www.washingtonpo…
  • @shiraovide Shira Ovide on x
    A genius glimpse by Drew Harwell into a new job: prompt engineers type coaxing words into AI chatbots to get machines to spit out optimal results. Idk, if we need AI sorcery to get machines to do what we want....isn't that bad? https://www.washingtonpost.com/ ...
  • @mountainherder @mountainherder on x
    There's something I can't quite articulate about this that reads like “the trick is to pray correctly so that God intercedes on your behalf” or trying to phrase a request to the djinn that can't be interpreted in a way that harms you. https://twitter.com/...
  • @washingtonpost @washingtonpost on x
    “Prompt engineers” are being hired for their skill in getting AI systems to produce exactly what they want. And they make pretty good money. https://www.washingtonpost.com/ ...