I think the current LLM trend is way overhyped. The thing these companies are doing, crucially, is extracting money from technologically unsophisticated companies and funnelling it into VC pockets. Silicon Valley bubbles are frequently incestuous. You make an HR platform and you sell it to other tech companies. You do sales analytics for tech companies. You make a bank that caters to only tech companies. The problem is that your customers all have the same risk profile as you - if interest rates go up and funding dries up you're all fucked at once. You're all just passing VC dollars back and forth.
The AI trend is luring in a wide range of non-tech businesses that aren't VC backed (ex. car dealerships). This is the golden goose for tech companies in a tough macro climate.
To be clear I think it's not adding real value for these buyers, but there's a tremendous amount of FOMO. The media has been complicit in selling AI as a panacea for small businesses in a tough climate, and the ultimate beneficiaries will be people grifting off AI hype.
Yeah, there is a kind of hysteria regarding AI and it looks like traditional business fear that if they don't use it somehow they will perish. But there are so many traps here, such as your customers not wanting interaction with your "AI chatbots", the issues of quality (vide the Sports Illustrated scandal), and the question of whether your niche needs any "AI" at all.
Yeah protobuf is a good IDL and encoding. Unfortunately gRPC makes some choices that make sense for internal RPCs in a large engineering org, but it's not good for external clients IMO.
There's an important distinction because a police "union" is in opposition to any other workers and unions. The police are the tools of capital, they break up strikes. The cops aren't going to join you in a sympathy strike.
VMs are a very interesting way to decompose a problem - you can implement the fundamental operations, test them thoroughly, and then generate bytecode and debug the bytecode implementation.
A couple years ago I thought about giving a presentation on a Golang VM I wrote to solve a simple problem, but it was seen as kind of a heterodox approach and I didn't want to deal with a million "umm acksully" questions about whether it was a real VM.
In Ontario my doctor sends prescriptions "electronically". In practice it's not clear whether this is like an email, or whether someone behind the scenes prints out the prescription and faxes it. Apparently the local clinic has a team that is solely responsible for faxing things on behalf of the doctors.
Here in europe it is a government database where all the clinics and pharmacies are connected to, so doctor essentially creates a record in the database.
When you visit pharmacy they ask for id and enter your id number system shows them all your active prescriptions and past ones as well, which sometimes helps when your prescription is not renewed for some reason they can give you a week supply while you sort it out.
That depends on the country. In France it varies by doctor, some will use Doctolib (a great third party private company that does appointment scheduling, video consultations and digital prescriptions) which allows you to have a digital prescription that you click on a button in the app/website to share with a specific pharmacy, and when you get there they just get your social security card and... then print out your prescription, and scan and print on it how and when was it fulfilled. Others just give you an old fashioned hand written note, or print an A4 sheet of paper.
Similar experience in Canada with our telecom cartel. They're legally required to allow other providers to use their lines but if you need support you're out of luck. We had multi-day outages routinely and a Bell tech couldn't find any issues. We finally switched to FTTH when it became available because we need reliable internet to work from home.
How would that even work? Are you looking at your surroundings or your work? It sounds like a nightmare version of people staring down at their phones while shuffling down the street.
They're automatically predicting the limit _and_ figuring out binpacking into hyperthreaded CPUs and NUMA cores. K8s just pushes your supplied values down to the kernel, which is exactly what they're saying is inefficient.
Tipping is still illegal if you know the recipient intends to trade. The person who asked him for advice trusted him with material non-public information, if he shares it and it gets traded on he's liable either way.
To be clear the person who asked for advice was not tipping because they had a reasonable expectation that he would obey securities law and not trade based on the information. They were asking for actual business advice.
The AI trend is luring in a wide range of non-tech businesses that aren't VC backed (ex. car dealerships). This is the golden goose for tech companies in a tough macro climate.
To be clear I think it's not adding real value for these buyers, but there's a tremendous amount of FOMO. The media has been complicit in selling AI as a panacea for small businesses in a tough climate, and the ultimate beneficiaries will be people grifting off AI hype.
reply