I could do very useful things with that machine. So it is not the end of the world if we have to go back to a world when you merely have thousands of times more memory for 4 times less money.

It could even be positive if it forces people to be more efficient writing code and wasting less resources.

It is one thing to go carving whenever you want, where you want because you have a good job outside it. Another totally different thing is spending all your time training. Most people will hate that.

Everybody wants to be a tennis player when they see one player raising the cup and earning millions. But a professional player spends most of her life doing extremely boring things. And only a very minority get enough money to live from the sport.

Going out for a walk alone to the beach or the mountains was forbidden. It was so ridiculous. And of course I went out anyway and it was essential for my health and sanity.

I have been working from home for a long time. It gives me freedom. I don't have to waste hours moving most of the time, and time with friends and family that I choose.

When I worked in an office I had to spend one hour moving in and one moving out each day.

This study is equivalent to the drugs study done on caged rats. The caged rats being humans during pandemic.

It was discovered later that if you let the rats freedom and the ability to socialise they did not get as anxious as when caged and they did not look for drugs for escaping their miserable lives.

I've been overweight in my life and I hate it. You know how I lost weight and kept if off? I ate healthier, I walk a lot (to the grocery store, for example). US grocery stores have all the same healthy foods as every other country. You can eat healthy at restaurants too, if you must.

It's just people don't. There's nothing inehrently bad about food in the US. We at eat too much and don't exericse (walk). Now, if you'll look at all the other countries joining us in this because it isn't a US thing anymore.

You've added all these extra explanations (corn syrup, antibiotics, growth hormones) but it really isn't that complicated. BTW, I'm sure if you ate konbini foods every day in Japan and didn't exercise enough you'd get fat, too. There's nothing different about that food than a lot of junk/fast food in the US, it's just the Japanese seem to walk a lot more and I don't know how many people only eat out of 7/11.

Except that sugar (or more often, corn syrup) is added to everything.

Just as a small example: most regular people don't have the time to bake their own bread or make their own spaghetti sauce from scratch, so they get both from the store. However, in the US, unlike most European countries (and probably the rest of the world), both have tons of added sugar and/or corn syrup. This is true for most foods in the US, and while it's certainly possible to eat healthy, it's a lot harder to do so here.

> and don't exericse (walk)

A big part of this, too, is that not only are most areas in the US not designed to be walking-accessible, but I daresay most are designed to be openly hostile to pedestrians. Added bonus: in the case where you do happen to live in a walking-friendly area, you still can't send your kids out to walk around, lest you invite visits from CPS.

That isn't to say people in the US can't be doing more than they are to get healthy, but arguing it's no more difficult to stay in shape in the US versus other parts of the world is at best self-delusion.

The saying is the US doesn’t have “health care” we have “sick care.” Preventing diseases? Not profitable. Giving people diseases then charging them for treatment? Profitable.

Fortunately the obesity epidemic has a cliff like tail end, as those people tend to eat themselves into an early grave and lack the resources to pay to prolong their participation in Planet Earth.

What inherently prevents you from cooking a tortilla or baking bread in the US vs Spain? Do frying pans and ovens work differently? Has the US outlawed paella?

Calories in > calories out.

The great news is that when you eat less, it brings down the relative costs of better food. Instead of pounds of low quality ground beef, 3 packs of Twinkies, and a case of beer; buy some higher quality organic chicken, or even better, some organic vegetables.

This is like a kid playing videogames instead of studying, you take the console away and force the kid in front of a book and the kid will spend most of his time looking at the wall and dreaming.

I am engineer with very deep programming background that have managed people, with real experience in the real world.

One of the best things about AIs is that you can test crazy ideas and create prototypes very fast. Only one in a hundred will work great in the real world, but you have to create the 100 before to know.

Creating the 100 before AI was extremely expensive, and took so much time.

For me it is liberating and gives me focus because I can spend so little time testing prototypes and spend real time in what is really important and works.

This is something I learned from game developers: If you are going to create a game, you spend a weekend testing the dynamics and the gameplay of your prototype to know if is is fun. You use boxes, no textures, no complex sounds of music.

Then if it works and is is so fun, you create the game! You can spend 2 years creating the game after that.

You don't spend two years doing a Game only to realise later that is not fun, and you either spend 3 more years or abandon it at this moment.

AI tools have put friction where it should be - by eliminating incidental friction. By incidental friction I mean, things that were really not ambiguous, but were made so due to lack of access to resources.

As an example, if i needed to navigate, I used a paper map. There was friction in pulling out a map, planning a route etc. This took time. With digital mapping apps this sort of incidental friction is not there.

Real friction is inherent ambiguity. For example, what product does the market need ? By eliminating incidental friction, AI allows us to focus on the smallest hard-problem where there is real-friction.

The wrinkle that needs to be added is that there are no truly universal rules as to what counts for incidental vs ambiguous friction - the definitions are relative to individual/project goals. I am working with some scientific instruments to map out chemical data, and 3d modeling is needed. I don’t particularly care about 3d modeling - it is incidental to me. The chemistry is the focal point. So the STL files are vibe coded so I can keep my focus on chemistry. But if I were working for the latest marvel movie, the reverse would be true. The actual chemistry would need to fit the script and the visual effect intended. To a scientist the visualization just needs to be good enough. To the film director, the world building physics and chemistry instead become the supporting actor. The challenge introduced by AI is that in ruthlessly eliminating incidental friction, you are being deliberate about what you choose not to learn. This is fine at a task level - but how many of us “found” our current expertise through incidental friction in the first place? I never wanted to do chemistry, I went to school for something else. But incidental friction led to discovery. That is my biggest worry, particularly for students and early career folks.

I’m much more able to navigate without the map than I was before I started this experiment - and it’s had the added benefit of giving me the ability to know what part of the 101 I’m on solely by the angle and shape of the nearest exit.

I hope eventually AI gets to a point where I can split the difference between ambiguous and incidental friction like this, although tbh I’m not really sure what that would look like.

People who speak those languages have an almost supernatural way of knowing where is north intuitively, because they HAD to do it to communicate.

Digital maps clearly solve the end-goal needs for most people. But like your father in law, there’s definitely a loss in that exchange.

Bearings are incredibly useful. I remember navigating myself and my partner out of a small town on vacation by the position of the sun. It was international so we didnt have internet at the time. Im never going to live that one down!

Agree, and I have the same worry, and so did the OP to the point that he considers canceling his AI subscription. Most of have to first get through the incidental friction to be able to arrive at expertise to address ambiguity.

That does happen on occasion, the commonly-cited example being Half-Life. How awesome would it have been if the Valve team hadn't had to waste so much time, money, and personal energy on their initial failed prototype?

Unfortunately most studios ship their failures, either because they don't realize they built something crappy or because the alternative is bankruptcy. A cynic would say that if AI can reduce the cost of experimentation, it will only result in more bad games, while an optimist would argue that it will result in more good games. I think we'll find that they're both right.

People do this all the time. It's such a common problem in startups that all of the books, courses, advisors, and everyone else with experience talks about finding product market fit early and shipping MVPs to validate the product.

It's the most common startup advice and people still ignore it and build unvalidated things for years anyway.

It's too easy to get started on your big idea and then switch to a rhythm of working on the next task without ever stopping to validate the big direction

Almost any reasonable idea that is not obviously bad (as in having some clearly insurmountable technical or market hurdles, etc) can be made to work.

What makes ideas work (and this is what separates those with a true entrepreneurs mindset from those just playing the lottery of "MVP"s) is the creation of what I call a viability field around them.

When a person decides to create an app, or some other product, or a certain business, they have not exhaustively considered all the possible things they could do and decided that this is the objectively most viable thing to pursue. They commit to a space they like ot be in, that feels like a good match for their skills and temperament, etc.

Then, within the general idea, the key I think is to begin creating a product, the development and marketing of which acts as a substrate for the learnings you need to succeed in that market. AI used judiciously can help you to cover ground more quickly, or at least be a bit more fearless about attempting that.

Basically you should build something that allows you to learn, to pivot, to adapt as necessary until you find product market fit. Rarely, you'd need to throw everything out, but even then I often argue it is matter of personal decision rather than any fundamental roadblock. If you you are really committed to an idea, and have developed it enough (not just a 1 week MVP) to deeply understand its context, there is almost always a tangent you can take that brings you to success without throwing out everything.

> Slopping out a 10,000 LOC untested Python/JS mess in 5 minutes helps nobody. The thought of this happening in every commercial environment simultaneously is horrifying.

You should def try to commercialize your slop and get feedback. No one cares about your tech stack or whether it's maintainable. Does it add value? You'll get a strong signal as to whether it does or not. And adding value, picking the right problem in the right domain, that's the hard part. You can always re-write or clean it up. Didn't Javascript start out as a weekend project? (maybe not a great example depending on how much you like JS)

People believe that you can only use LLMs for sloppy programming. But you can also use it for writing ten times more code of Swiss cheese model tests, and domain specific languages.

You write ten times more code than necessary and all that extra code is testing. Projects like SqlLite do that because they need to be perfect.

Before LLMs we had to use engineers for that and it was a painful and repetitive work, and they were always late and made much more mistakes than LLMs, specially because it was dull and tedious for great engineers to spend their time into.

Now we write tests and when all test pass we write new test for checking the tests.

We divide each complex problem in small subproblems and we warrantee each of them by formal means. We have multiple ways of solving the same problem, usually with one brute force solution that is simple and warranted to work but inefficient, and we can use it to compare with more efficient methods.

Before machines could do that, people doing that were burned down and exhausted, and always leaved pending work to complete.

Care to expand on how is this accomplished?

So low in fact that the majority of the recent "accidents" look like suicides from the pilots. The pilots know exactly what they are doing when crashing the planes.

Please listen to Isaac Moreno Gallo. 99% of the Roman Roads had no big stones on it. Only near big cities you have the stone pavement, basically in the cemetery that was outside town alongside the road.

People with very little idea about engineering wrote the textbooks of the past and some of the wrong ideas are transmitted even today.

For people that have been programming for 20 years, AI is not an issue, on the contrary, it can work for you as you become older. They have trained enough.

For young people, it can destroy them if they stop practicing hours every day, outsourcing all their work to the ai, as they will start losing contact with the reality they are controlling.

But that depends on the personality of people. One of the best uses of LLMs is creating trainers for practicing what you need. For example, I have created programs for practicing my Japanese and Mandarin pronunciation, for practicing chords and note recognition with MIDI, and playing of sheet music, practicing IPA, my handwriting...

Usually I buy some software that does more or less what I want, but them I realise that I need a specific feature the software does not have. So I create the trainer for personalization.

It will be impossible for me to do that without LLMs. There is no time in 5 lifetimes to do that by hand.

I have created 20 utils or so with the help of Claude, in order to practice multiple things like reading sheet music, or rhythms, or different scales. I never expected it to be that useful as my new Yamaha was bought before Claude existed, and having a cable that just works is so great.

I have spent way less effort doing all my utils than this man into just connecting its machine.

Before using it with Claude I used them a lot with Synthesia and GarageBand, but with Claude is like having a personal trainer.