It's how venture capital works: the theory is that you invest in a dozen pie-in-the-sky pipedreams, almost all of them fail, and the one that doesn't pays for the rest. If you're not failing most of the time, you're not taking on enough risk.
When first encountering these ecosystems and looking at the various pieces they contain, one may repeatedly ask: "why didn't they just use <off-the-shelf solution> for this problem instead of writing this component/subsystem from scratch"?
The answer is often that the battle-hardened mature off-the-shelf solution did not exist at the time the code was written. You're doing software archaeology.
For the love of all that is good, clean up the AI slop before you parade it in front of people. Can't you see the misspellings, misattributions and duplicated labels in the images? If you couldn't be bothered to spare it even a passing glance, why should I bother? I, too, can prompt an LLM; so what, exactly, is the point of you?
Note that those are total comp numbers. Something like 60% of that is dependent on performance review, which is fed in large part by automatically collected metrics like number of commits, number of code reviews ticked off, the LLM token usage leaderboards etc; creating all the perverse incentives you'd expect.
So you can rely on income to precisely the extent you are willing to destroy your soul.
So, Meta are certainly happy to take advertiser money (except when they aren't, apparently), and demonstrably triple-digit-billion-dollars worth of people are happy to pay them, so something somewhere must be working, but I still don't understand...
...who the hell buys anything from Facebook ads? I never have, no-one I know ever has. Is my bubble seriously that strong?
I admit, I do click on Facebook ads every so often. This likely means Meta gets paid, and also some number somewhere goes up which means I get shown more of the kind of things I click on. This is how I end up seeing adverts for hi-vis vests for poultry, radioactive pendants (sturdy titanium, reliably glows for 25 years!), tungsten cubes and so on; because I see them and I think "what the actual..." and I click, because you can comment on Facebook adverts and maybe there is some kind of sanity to be found; and, as expected, the comments are full of confused people saying "what the actual..." none of whom are any more likely to buy the product than me.
Has anyone here, but especially of the people considering using Facebook to advertise their product/service, ever bought anything from a Facebook ad? What was it?
The job contracts include clauses about not posting things that paint the employer in a bad light on social media. No idea how enforceable that stuff is in any given jurisdiction, but really, what individual can afford to find out?
Your “and then” is doing a lot of work there. The steps between may or may not include some form of “learn to understand humans”, but you can’t just hide them behind “and then” if what we are doing is claiming some particular thing is not in the list.
Through training on human text, we are building implicitly in the weights a statistical model of what humans might write in response when presented with arbitrary pieces of text. It turns out that we can make these incredibly accurate.
If building an accurate internal model of something then using it to predict that thing’s behaviour is different to gaining understanding of that thing, we will need to pin down exactly what “understanding” means, or we are forever doomed to talk at cross purposes.
Turing complete systems can be built out of matrix multiplications, out of attention, out of key/value lookups. The Chinese room is Turing complete. By claiming it cannot understand things because it is built out of components computing devices can be built out of, we are claiming no computer can because no computer can. This is a very bold claim indeed, and also we’re assuming the conclusion! The claim is no more convincing than “brains cannot understand things because they are made out of neurons”. The system may or may not have some particular properties, but we have to do more work than just gesturing at the components the system is made of when making claims about it; the alternative is, at best, a world where we prove too much and conclude that humans, too, are not conscious.
For starters, we need to pin down the terms under discussion enough that they don’t just mean whatever we need them to in the moment.
Does anyone remember the early 2000s joke virus emails? The ones that are variations on "This is a <outgroup> computer virus. As we don't have software engineers to write the code to do this automatically, please kindly forward this email to everyone in your address book then format your hard drive."
This is exactly as much malware as those were.
Please, for the love of all that is good, can we just try not to build and defend a world where, on encountering text like that, /your computer immediately follows the instructions/? Can we just all agree that such a world would be bad for everyone involved and using an LLM that risks doing this, with no container or guardrails, is at least as problematic as running an unpatched open email relay was back then?
Flagship chip yields are generally less than 50%. Over half the chips coming out are dead on arrival, and never leave the fab. Imagine if rockets had that sort of failure rate.
Genuinely opening up is a mistake. The incentives for these clearly mean that they actually select for candidates who are capable of glibly blagging their way through an extended conversation without saying anything inconsistent, weird, compromising or of substance.
This isn’t usually a required engineering skill. I’m guessing the interview was designed for salespeople and/or middle management.
You are the google search engine pre-2010, well before Google lost their "don't be evil" motto, made the first results page favour sponsors and added AI overview. You respond to a search query with a list of https:// URLs, each accompanied by a representative quote from the destination page that demonstrates the link's relevance to the query, and nothing else. The query is: <insert your query here>
We live in the dystopia we deserve. We have built it with our own hands and it is here to stay.
Turns out that when you're implementing network applications, the set of things that could happen also depends on what the script kiddie on the other side of the globe feels like this morning.
Some would prefer less excitement than this.
C code should be more predictable and easier to reason about than using a macro assembler. To the extent it is not, the language has failed.
“Implementation defined behaviour”: compiler author chooses, and documents the choice.
A lot of UB should be implementation defined behaviour instead; this would much better match programmers’ intuitions as they reason about their code - you can even see it in the comments of this post: it’s always things like “this hardware supports / doesn’t support unaligned accesses”, it’s never nasal demons.