Because LLMs don't understand anything. That's the tech. They can only predict what they have been trained with and fail daily at the most basic tasks. Granted they can do amazing things, no question there. But they are not "smart".
For example, it seems that even at Fable scale, simple concepts like the passage of time or (gasp) timezones elude them. I live in UTC+10 and with any RFC8339 data LLMs are constantly confused - is it Sunday the 10th or Sunday the 9th, etc. I have tried many solutions for this and every time it finds a way to get it wrong.
How do you square that then? They can do amazing things, but they're also not smart? Do you think its possible to solve Erdos problems without any "smarts"? Can you do it without even understanding mathematics?
I find it very hard to hold the idea that LLMs don't understand anything. They can explain concepts, translate them, simplify them and implement them in code. From the outside, LLMs seem to understands most concepts better than most humans do. Do you understand anything? Couldn't I make the same argument? How would you prove that you understand what a for loop is, or that you know what calculus is? I assume you'd demonstrate your knowledge by using a for loop in a program, or explain calculus back to me. But LLMs can do that too.
> For example, it seems that even at Fable scale, simple concepts like the passage of time or (gasp) timezones elude them.
Funny example, because lots of human struggle with this too. The number of meetings I've had with people in the US! "Lets meet on thursday morning australia time!". Only, they actually meant thursday night US time, which is friday morning australia time. "Oooh that's so weird! Its the next day for you!". ...... Yes, I know.
I think LLMs are just a different kind of intelligence than humans. They're better at some things than us, and worse than others. They can find latent security vulnerabilities in the linux kernel, but struggle to count the Rs in strawberry. They're not as smart as humans in many ways. But we're not as smart as LLMs in plenty of ways too. I didn't find those linux bugs.
> The Chinese cars are taking over here: it’s a product people want at a price they like
If you can hazard a guess, which make and model is the “Tesla killer” for EVs, if such a car exists.
I frequently suggest to folks in the US (where I’m from) that BYDs in the US would change the competitive landscape, but I can’t reliably point to a make/model or two that they can check out online.
I have recently moved from US to Australia and it is 100% clear to me the US automakers will get absolutely crushed by the Chinese companies if/when they are able to access US market.
Especially EVs and PHEVs. This place is awash with them, they are cars people want at the right price.
This has been a concern for me too. But the agent is just a statsd receiver with some extra magic, so this seems like a thing that could be solved with the collector sending traffic to an agent rather than the HTTP APIs?
I looked at the OTel DD stuff and did not see any support for this, fwiw, maybe it doesn't work b/c the agent expects more context from the pod (e.g. app and label?)
Yeah, the DD agent and the otel-collector DD exporter actually use the same code paths for the most part. The relevant difference tends to be in metrics, where the official path involves the DD agent doing collection directly, for example, collecting redis metrics by giving the agent your redis database hostname and creds. It can then pack those into the specific shape that DD knows about and they get sent with the right name, values, etc so that DD calls them regular metrics.
If you instead went the more flexible route of using many of the de-facto standard prometheus exporters like the one for redis, or built-in prometheus metrics from something like istio, and forward those to your agent or configure your agent to poll those prometheus metrics, it won't do any reshaping (which I can see the arguments for, kinda, knowing a bit about their backend) and they just end up in the DD backend as custom metrics, and charge you at $0.10/mo per 100 time series. If you've used prometheus before for any realistic deployments with enrichment etc, you can probably see this gets expensive ridiculously fast.
What I wish they'd do instead is have some form of adapter from those de facto standards, so I can still collect metrics 99% my own way, in a portable fashion, and then add DD as my backend without ending up as custom everything, costing significantly more.
Inngest has built in concurrency controls and rate limiting to prevent systems from being overloaded, allowing the user to have the same controls of a traditional worker, but just a simple config option.
Inngest was designed to be a solution that can replace traditional queuing systems and event driven systems. Originally, it was built with the idea to handle complex flows that need to be durable, e.g. healthcare workflows that have time-based follow ups and legal compliance tasks that need to be executed in a specific order depending on patient confirmation or other actions.
We've seen users build all sorts of things: automate infrastructure based on events, building AI Agents around LLMs, perform vulnerability scans across thousands of packages, building scheduling products, and e-commerce data pipelines. We're seeing new use cases each week.
This seems very similar to Secret [1]? IIRC there were some competitive apps at the time, YikYak maybe?
Secret sounded scandalous, and maybe that was the goal, but it actually did a really great job of anonymous Q&A, and was full of posts from people saying "I have this private/odd/embarrassing situation what should I do", and more often then not the thread contained some really insightful and thoughtful stuff.
I was disappointed when it shut down, is the idea of this to be similar?
OK well then, we'd better not do anything about it! <s/>
I don't think this is strictly true. Oil production is not fully fungible due to all kinds of issues including sanctions and production capacity.
But this is also a strawman - this isn't saying "we should shut them all down" or even begging that question.
At a minimum it's saying "there were people and companies who intentionally chose to mislead the public on a potentially existential issue - perhaps there should be some reckoning for that?"
I suspect we had one chance to do it right, but the organised and widespread suppression and misinformation campaigns of these companies basically made that impossible.
Now we're faced with a range of unpalatable choices and a population fiercely divided about what to do.
This is a common but erroneous thinking. Man influenced climate change is an ongoing process.
It would have been great to start doing something thirty years ago to avoid some of what’s happening now but we didn’t. Now we still have to do something or it will get even worse.
This is not a throw your arms in the air and wait for the end while screaming situation.
Not advocating throwing my hands in the air, but the deliberate delay has narrowed our choices at a rate faster than technology has come to the rescue.
There is going to be a lot of complaints as choices keep getting removed and decisions forced upon us by the escalating crisis. If you are a freedom loving libertarian, you are not going to enjoy the next few decades. Which would be deliciously ironic if it wasn't such a monstrous tragedy.
What we are starting to meaningfully see, at least in Europe, is regulations which actually have teeth regarding where investments go and what companies can do regarding emissions and sustainability.
It means some things are going to either get more expensive or downright disappear but I wouldn’t call that a significant curb on freedom.
It’s more akin to not being able to use lead in plumbing than living under an authoritarian regime. Sure there will be less meat on the menu but, well, c’est la vie.
Hard disagree. We are well beyond the point where farting around on the edges of consumer behaviour can ever make a dent in reversing climate change.
Blaming consumers buys into the big emitters disinformation campaigns. Its not our fault.
The big emitters need to be curbed hard and fast. Our governments so far have been unwilling, complicit or unable to do so.
Its going to require a mass, popular movement to shut down coal mining worldwide, to restrict air travel, to localise food production and reinvigorate ground based mass transit. Nobody can make that happen as individuals at the checkout.
European regulations are not oriented towards consumers nor are they “farting around”.
These are far reaching regulations on investments. They impact literally everything. It’s going to be hard to do something if you can’t get money to do it.
But these are not hard choices or libertarian hell. That’s just a bit more regulations.
Sure, mining coal won’t be profitable anymore, people won’t be able to buy diesel cars and both flying and meat will become more expensive. It’s definitely a shift in how people live and probably a downgrade in some aspect regarding quality of life. But it’s a relatively painless one.
For example, it seems that even at Fable scale, simple concepts like the passage of time or (gasp) timezones elude them. I live in UTC+10 and with any RFC8339 data LLMs are constantly confused - is it Sunday the 10th or Sunday the 9th, etc. I have tried many solutions for this and every time it finds a way to get it wrong.