To me placing the join predicates immediately after the tables is more readable as I don’t have to switch between looking at the from and where clauses to figure out the columns on which the tables are joined.
If drone armies are vastly more powerful than human armies, the latter become insignificant much like unarmed civilians are irrelevant in a war today. If Ukraine loses its army, it will surrender. If an army of the future loses its drones it will be impossible to oppose the winner’s drones and would surrender as well.
Still had to atom bomb the Japanese for them to capitulate. Sometimes defenders ready to fight with stick and stones. At the end of the day, still need to slaughter a bunch of combatants and civilians to extract human cost to prime surrender. Drones may make that more expedient by making the slaughter competely lopsided. But peoples tend not to capitulate until they've suffered severe loss in material and lives. IMO what's more likely are fully automate drones that guns down wedding parties without human operators on winning side developing PTSD.
They make an (easily made) mistake on that page: the encrypted version of FTP is not SFTP, but FTPS. SFTP is an entirely different protocol based on SSH.
Yes, and it was a disaster. The Houthis could aim their missiles at Saudi oil infrastructure again. Much simpler to let America handle both finding a solution and taking the fire.
Those are the same conflict though. The saudis are fighting the Houthi’s in the Yemeni civil war. The same group that’s attacking the ships in the Red Sea.
But I feel well entertained watching them fall. Like using them and experimenting around.
But it also shows the road ahead quite clear. For example were is the money coming from? From millions of people paying for GitHub copilot for example.
How is it sold? Per webui, API and cloud providers.
Digital twin will also play a huge role in this as a bridge between AGI <> real world.
Why obviously? It might as well be a passing fad, and all those uses oof ChatGPT might turn out to be a temporary amusement rather than real and lasting improvement of people's workflows. It's interesting how something comes out, and then all of a sudden so many people are immediately absolutely certain that it will have a massive impact at many levels, even before we've seen any meaningful ROI. To me it is a bit absurd and somewhat annoying. Of course, the tech is cool, it does have some amazing uses, but to forecast growth to trillions of dollars over the next few years and massive job losses seems premature and fuelled by relentless promotion, not only by the likes of OpenAI, but also by all the investment-hungry tech businesses - large and small.
For many people it is already a lasting improvement. It simply saves us so much work that we can improve in many other areas. And that is improving too; we have a slew of internal tools build with several LLMs, including the openai ones, that effectively replaced full time employees. The entire process of transforming arbitrary json or xml to another json with a required knowledge of the field semantics is now done quite perfectly using LLMs. And that is a lot of the work we do. Creating json schema’s based on a pdf, text, arcane line feed format etc is also now seconds vs hours. Debugging previous (and we have 10000s of these) transforms is also automated and simply, measurable, more accurate and faster than humans. And it was boring work so we can focus on other things.
Well it's already here if you care to look. So much stuff is already generated by AI, I use it, many non-technical people I know use it. They didn't have to be taught how it works, it's very accessible, it just works and it can save time and money.
Now there are plenty of challenges to overcome of course, but I have no doubts that something that useful on day one is going to have a big impact once we really understand how to integrate it to various products.
Honestly? It's the tech people who have this weird blinkered view on it. There's a zoomer clique that's mad about it, but beyond that, just watch the NYT OpEd page to see how normies are engaging
I don't think it's a passing fad, but the legitimate and lasting use cases are lost among the hype and bullshit.
There WILL be job losses but it's not going to be the kind of people who hang out on HN. I can't think of any reason why you wouldn't have an AI taking orders at the drive through, handling customer service calls / tech support, etc. Any job that consists mostly of having the same simple, repetitive conversations is going to eventually be cheaper to have a computer do.
Well, if you have multiple queries, you'd still have to write the cte multiple times. You could also use views for the same thing, and those are much more reusable.
Yeah. In many cases I've had to use CTEs like this in BI tools. When you're experimenting with datasets for dashboards it's much faster to work with CTEs than to try to make production DB changes.
I agree, I'm often using CTEs to organize sub-result-sets because asking for a production database change / view change requires a ticket, and it's just way faster to iterate in a few CTEs in a single query.
If, eventually, a few CTE distillations become common enough, then yeah, that's grounds for a request to basically shove that CTE into a view so other people can use it.
A small bit of advice from me: consider the weight of the binoculars when buying them. They can be too heavy to hold for more than a minute - especially large ones made for astronomy. You will either want a tripod (which is inconvenient to carry around and set up), or you’d want the binoculars to be smaller in size.
For any practical viewing with binoculars you do need a good tripod - otherwise instead of stars you’ll see shaky lines. Unlike daytime viewing your eye/brain doesn’t eliminate small shakes, or maybe it’s just harder to stabilize binoculars when looking up.
While working for one of the largest consultancies, I had to interview an offshore candidate for an onshore position in my team. It was a database development job and I asked the typical SQL questions, data modelling, etc - all over Skype with no camera. The guy on the other side seemed very knowledgeable. Amazingly high level for the job actually, and better than what his CV suggested. Being an internal candidate, I did not suspect anything and recommended him for the job. A month or so after that, management flew him in and I got to meet him in person. To my surprise, his language skills and much less his technical skills were anywhere near what I experienced at the interview. He was very keen to be a manager, with very very poor SQL skills - even in theory. A friend of mine later told me that it is a common practice in the offshore dev center to let the good devs interview for positions and then to send another person to do the job. They apparently did that to clients quite often and it was not surprising to him that I got the same treatment - even though it was an internal assignment.
I mean, this is not incredibly different from consultancies sending their crack A-team to pitch a project and then a bunch of freshly-recruited graduates to actually build it. All big firms do it. I can see some particularly entrepreneurial fellow overseas experiencing it once and then going "well, if they can get away with it, why couldn't we...?"
