commonturtle

The simple answer: those companies aren't evil. They're not selling your data, or bringing the financial system down, or emptying the wallets of the poor. They're not god's gift to mankind or finding the cure to pancreatic cancer either. They're just fine.

simply because that's also not how humans write novels.

Planes don't fly by flapping their wings either. It's totally possible we will construct a super capable artificial system that only vaguely resembles the human brain.

Personally I think if we managed to create a 10 times more sophisticated GPT-3 (instead of 2048 tokens it looked back 10k+ tokens) we might get pretty close to a novel writing AI.

Really curious to see what comes out of this.

Very inspiring. And here I am at the age of 28, wondering if I'm too old to get a PhD.

Yup that's a good recommendation. I've read it and some of the AI Safety work that a small portion of the AI community is working on. At the moment there seems no reason to believe that we can solve this.

This is simultaneously amazing and depressing, like watching someone set off a hydrogen bomb for the first time and marveling at the mushroom cloud it creates.

I really find it hard to understand why people are optimistic about the impact AI will have on our future.

The pace of improvement in AI has been really fast over the last two decades, and I don't feel like it's a good thing. Compare the best text generator models from 10 years ago with GPT-3. Now do the same for image generators. Now project these improvements 20 years into the future. The amount of investment this work is getting grows with every such breakthrough. It seems likely to me we will figure out general-purpose human-level AI in a few decades.

And what then? There are so many ways this could turn into a dystopian future.

Imagine for example huge mostly-ML operated drone armies, tens of millions strong, that only need a small number of humans to supervise them. Terrified yet? What happens to democracy when power doesn't need to flow through a large number of people? When a dozen people and a few million armed drones can oppress a hundred million people?

If there's even a 5% chance of such an outcome (personally I think it's higher), then we should be taking it seriously.

Human's want health insurance and 40 hours work-weeks. The super-smart AGI that will exist 20 years from now won't.

You'll still need humans to make anything novel or interesting, and companies will still need to hire engineers to work on valuable problems that are unique to their business.

Give it 10 years :) GPT-10 will probably be able to replace a sizeable proportion of today's programmers. What will GPT-20 be able to do?

Yes he's a little more free-market than I am. But I like that he's at least willing to consider the other side's position and invites guests who don't agree with his positions often.

Machiavelli's The Prince :)

EconTalk is my current favorite. Around 700 episodes since 2006 on a variety of subjects: education, healthcare, public debt, etc. The host dives quite deep into each subject and asks the interviewees tough questions. I always feel like I've learnt something from each episode.

Apps and the associated infrastructure (driver background checks, payments processing, customer support, etc.) are much harder to build than they look. I'm willing to bet against a worker-owned app taking >= 10% of Uber and Lyft market-share in >= 10 mid-size American cities.

This view on bio-ethicists reflects my views on ethics as an academic field in general: https://marginalrevolution.com/marginalrevolution/2020/12/fr...

Some quotes:

I realized quickly that the incentives are all wrong if what we want is people who will think hard about humanity’s pressing ethical dilemmas and who will suggest intuitively appealing solutions.

Since almost all ethicists are academics, they have to publish, and in order to publish you have to be novel, and since the basic principles of ethics are little changed for millennia the incentives to do thorough homework on the basis of principles which are widely understood and accepted is not great.

Furthermore, if you decide to be a utilitarian, then basically all ethical issues will boil down to cost/benefit analyses which you have to outsource to technocrats, so your unique expertise as an ethicist will be worth little.

German tech salaries are quite low compared to US and UK last time I checked.

That's a really fun form factor for touring a city. If they start offering these to tourists as rentals, especially in cities that have narrow streets (like most of Italy), I would happily use them.

It's quite a low wage, but not 5 rupees an hour: 8000 / 160 = 50. 50 rupees an hour is probably a decent wage in a small town / city in India, but would be a tough wage to live on in Bangalore.

These are little optimizations though. I see asking taxi drivers to pass The Knowledge as the same as asking students to use logarithm tables or memorize digits of PI: yes there is some marginal benefit to doing it, but honestly there are so much more valuable things one can do with their time and energy.

Nice overview of how these small software projects are sold. I'm also curious to read about the other side of the transaction; who buys these small projects and how do they monetize them further?

Selection pressure / natural selection. Any new strain that is more transmissible will transmit faster than other strains, and therefore grow in prevalence compared to other strains.

It doesn't matter these days. Google Maps just works, I've driven in London dozens of times with absolutely no issues.

There is zero reason to continue to have such a complicated test for taxi drivers when satnavs are so effective, all it does is raise costs for riders and make it harder for people to become drivers.

I've been trying to find a camera cover that will allow me to cover both the cameras without obstructing the infrared light and the microphone. No luck yet.

Amazing. I found a documentary about the aircraft on youtube: https://www.youtube.com/watch?v=qfENcznI8e0. The flight across the English channel starts at 41:33.

We have two night trains in the UK. One of them is the Riviera express that goes west from London to Penzance in Cornwall. I’ve traveled on it, it was an expensive but convenient experience.

The second is the Caledonian Sleeper that goes from London to Inverness in the Scottish highlands. This one is ridiculously expensive. I’ve never seen tickets that cost less than £300 for a lie down seat, which is 8-10x a typical airfare 3-4x a typical seated train fare. Probably will never travel on it but I’ve always wondered why it’s so expensive. If demand is so high, the why not increase supply by adding an extra train?

Wonderful idea. Seems very ML focused at the moment, would love to see other fields here too.

I can do at most 4 hours of high intensity focused work in a day. But usually administrative work at my job takes up so much time (interviews, meetings, standups, etc.), that I don't have to worry about that limit anyway.

It takes time to make the economics favorable by incrementally improving the technology. Once indoor vertical farming does become favorable, it will have huge benefits over outsourcing food production to Africa or elsewhere.

National security for one; it times of war no one can starve your population into submission by cutting off shipments of food into your country. Your crops are also more resilient to pestilence and disease if grown in a controlled indoor environment.

All this makes it a worthwhile line of exploration, no?

I've been working from a different city every month for the last few months. It's a good way to do slow travel.

I've seen this discussion come up on HN a few times so I think people may find Tyler Cowen's talk on this subject interesting.

Some interesting bits of information:

- People do work less to some extent, but the reduction in working hours is concentrated in teen years and old age (65+).

- People seem to like work, will continue their jobs even if they win the lottery.

- Rich people (top 1%) now work longer hours than poorer people.

I also struggled with this transition a bit so here are some of the things I learned that help me get to L5:

1/ The key thing for getting to L5 is to take on large problems (usually 3-6 month long efforts), figure out how to solve them, and ship them successfully. You can do this by leading a small group of people (2-4 engs), but also by working as an individual, so I don't think leadership skills in that sense are required for L5.

2/ Usually managers are quite good at coaching for this transition (things get harder going from L5 => L6 and further on). Try to find a team that is seeing some growth and with a manager you enjoy talking to. Managers with some experience (2-3 years) are better at figuring out what gaps you need to fill to get promoted.

3/ Also try to figure out what type of team you will be successful in. For introverted people often product teams are not a great fit as they require tons of coordination to ship anything: you might have to work with a designer, a content strategist, a product manager, and a data scientist for a typical L5-scope project. Infrastructure / backend teams can offer better opportunities for people who want to grow as individual contributors. I know many senior engineers (L6+) in infra teams who aren't tech leads leading large teams -- they are simply strong engineers who can solve complex problems on their own.

I don't understand how this could happen in places that have good reviews. I only consider places with a minimum of 10 reviews and avg score > 4 stars. Never had a bad experience.