Login

Recent searches

No recent searches

Search options

Not available on writing.exchange.
writing.exchange is one of the many independent Mastodon servers you can use to participate in the fediverse.
A small, intentional community for poets, authors, and every kind of writer.

Administered by:

Server stats:

323
active users

writing.exchange: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#automation

10 posts10 participants0 posts today
Public
Miguel Afonso Caetano

"The challenge, then, isn’t just understanding where A.I. is headed—it’s shaping its direction before the choices narrow. As an example of A.I.’s potential to play a socially productive role, Autor pointed to health care, now the largest employment sector in the U.S. If nurse practitioners were supported by well-designed A.I. systems, he said, they could take on a broader range of diagnostic and treatment responsibilities, easing the country’s shortage of M.D.s and lowering health-care costs. Similar opportunities exist in other fields, such as education and law, he argued. “The problem in the economy right now is that much of the most valuable work involves expert decision-making, monopolized by highly educated professionals who aren’t necessarily becoming more productive,” he said. “The result is that everyone pays a lot for education, health care, legal services, and design work. That’s fine for those of us providing these services—we pay high prices, but we also earn high wages. But many people only consume these services. They’re on the losing end.”

If A.I. were designed to augment human expertise rather than replace it, it could promote broader economic gains and reduce inequality by providing opportunities for middle-skill work, Autor said. His great concern, however, is that A.I. is not being developed with this goal in mind. Instead of designing systems that empower human workers in real-world environments—such as urgent-care centers—A.I. developers focus on optimizing performance against narrowly defined data sets."

newyorker.com/magazine/2025/04

The New Yorker · How to Survive the A.I. RevolutionBy John Cassidy
#AI#GenerativeAI#Automation
Public
Miguel Afonso Caetano

"Dwarkesh Patel: I want to better understand how you think about that broader transformation. Before we do, the other really interesting part of your worldview is that you have longer timelines to get to AGI than most of the people in San Francisco who think about AI. When do you expect a drop-in remote worker replacement?

Ege Erdil: Maybe for me, that would be around 2045.

Dwarkesh Patel: Wow. Wait, and you?

Tamay Besiroglu: Again, I’m a little bit more bullish. I mean, it depends what you mean by “drop in remote worker“ and whether it’s able to do literally everything that can be done remotely, or do most things.

Ege Erdil: I’m saying literally everything.

Tamay Besiroglu: For literally everything. Just shade Ege’s predictions by five years or by 20% or something.

Dwarkesh Patel: Why? Because we’ve seen so much progress over even the last few years. We’ve gone from Chat GPT two years ago to now we have models that can literally do reasoning, are better coders than me, and I studied software engineering in college. I mean, I did become a podcaster, I’m not saying I’m the best coder in the world.

But if you made this much progress in the last two years, why would it take another 30 to get to full automation of remote work?

Ege Erdil: So I think that a lot of people have this intuition that progress has been very fast. They look at the trend lines and just extrapolate; obviously, it’s going to happen in, I don’t know, 2027 or 2030 or whatever. They’re just very bullish. And obviously, that’s not a thing you can literally do.

There isn’t a trend you can literally extrapolate of “when do we get to full automation?”. Because if you look at the fraction of the economy that is actually automated by AI, it’s very small. So if you just extrapolate that trend, which is something, say, Robin Hanson likes to do, you’re going to say, “well, it’s going to take centuries” or something."

dwarkesh.com/p/ege-tamay
#AI #LLM #Reasoning #Chatbots #AGI #Automation #Productivity

Dwarkesh Podcast · AGI is Still 30 Years Away — Ege Erdil & Tamay BesirogluBy Dwarkesh Patel
Public
Miguel Afonso Caetano

That's Late Stage Capitalism - China Style ->

"China still dominates shoe production, but its share of global exports has slipped by 10 percentage points over the past decade, with much of that going to rival hubs like Vietnam and Indonesia, according to the Yearbook.

Zhou’s plant now employs fewer than 20 workers. “The future is bleak and hopeless if we continue like this,” says Zhou from his showroom in a mostly empty wholesale market in a Guangzhou suburb focused on international trade. “It would be difficult to return to how it was before.”

Factories across China at the low-end of manufacturing are facing the same dilemma — either they invest in automation that shrinks the number of jobs, or they slowly wither away.

The result, in the view of researchers and economists, is a painful shift away from low-cost, labour-intensive production that could leave millions of older, lower skilled workers in the lurch.

Analysis of 12 labour-intensive manufacturing industries between 2011 and 2019 by academics at Changzhou University, Yancheng Teachers University and Henan University found that average employment shrank by roughly 14 per cent, or nearly 4mn roles, between 2011 and 2019. Roles in the textile industry shrank 40 per cent over the period.

An FT analysis of the same 12 sectors between 2019 and 2023 found a further decline of 3.4mn jobs."

ft.com/content/7640fe64-006a-4

#China#Manufacturing#ShoeProduction
Public
Yuna

Your logs are lying to you - metrics are meaner and better.

Everyone loves logs… until the incident postmortem reads like bad fan fiction.
Most teams start with expensive log aggregation, full-text searching their way into oblivion. So much noise. So little signal. And still, no clue what actually happened. Why? Because writing meaningful logs is a lost art.
Logs are like candles, nice for mood lighting, useless in a house fire.

If you need traces to understand your system, congratulations: you're already in hell.

Let me introduce my favourite method: real-time, metric-driven user simulation aka "Overwatch".

Here's how you do it:

🧪 Set up a service that runs real end-to-end user workflows 24/7. Use Cypress, Playwright, Selenium… your poison of choice.
📊 Every action creates a timed metric tagged with the user workflow and action.
🧠 Now you know exactly what a user did before everything went up in flames.

Use Grafana + InfluxDB (or other tools you already use) to build dashboards that actually tell stories:

* How fast are user workflows?
* Which steps are breaking, and how often?
* What's slower today than yesterday?
* Who's affected, and where?

🎯 Alerts now mean something.
🚨 Incidents become surgical strikes, not scavenger hunts.
⚙️ Bonus: run the same system on every test environment and detect regressions before deployment. And if you made it reusable, you can even run the service to do load tests.

No need to buy overpriced tools. Just build a small service like you already do, except this one might save your soul.

And yes, transform logs into metrics where possible. Just hash your PII data and move on.

Stop guessing. Start observing.
Metrics > Logs. Always.

#Observability#MetricsOverLogs#DevOps
ExploreLive feeds
About