Dijkstra's personal language design (GCL) actually used anything-based indexing -- you specify the lower bound explicitly when you declare the array, so even if indexing starting from some runtime value k (e.g. k=7, k=b.lower-bound for some existing array b, etc.) is most natural, you provide that as the lower bound. Then a(i) would be translated to e.g. a[i-k] in compilation.
Getting Inform7 to work well probably depended on 'text adventure'/IF culture, since the text adventures themselves rely on this kind of 'restricted natural language' interface to navigate. Conversely, IF developers built up a lot of intuition on how to make pseudo-natural-language input feel natural while keeping the restrictions obvious.
> The disclosed AI use was apparently limited to inventing the name of a fictional disease. The organizers still counted it among the year’s AI-assisted winners.
This one was interesting to see, because cases like this make the difference between luddite exclusion (anything that transitively comes from 'AI' is spiritually-tainted bits) and accurate disclosure (LLM output is distinguished from human product). Either policy is a valid choice, but lumping the two together as equally 'AI-assisted' tends to confuse rather than help.
This is similar to the 'no gray areas' problem with compiling a list such as https://codeberg.org/ethical-foss/ai-free-foss-alternatives repo and needing a complex system of tags to avoid mixing projects that have measured/conservative stances short of full exclusion, with projects that have gone full slop-cannon.
> Only developers think that in most cases, the hard part is code. Its never been.
The correct way to think about this:
The hard part is ensuring the product does not exhibit undesired behaviours.
Writing code is one possible tool for ensuring this.
Replacing writing code with LLMs only works to the extent that LLMs accurately predict your definition of 'undesired'.
The more unique and innovative your product, the more likely the LLM prediction is to be inaccurate.
Totally agree. But most products aren't innovative. Currently in IT, the most innovative product its actually the LLM. You know who else sucks at writing greenfield code? Hunans.
There are... multiple blog posts online now about how to use freely available data and modest amounts of compute to train a custom GPT-2-sized model from scratch. It would be quite a policing effort to prevent.
Clearly Apple's decision is based on paternalism, which most people don't appreciate when it's combined with government power. Compare:
"Clearly the ability to eliminate opposing viewpoints was a huge part of the Chinese government's decision, but was far from the only one. Stability and economic prosperity are also very important. My grandma can obey simple laws. But telling her to go to research contradictory political perspectives and decide the correct course of action is a completely different story. Not to mention how many people will just vote for the candidate that pushes their emotional buttons, and read whatever crap on Facebook and click next, next, next until they've elected Donald Trump."
For sure, and I don't personally recommend that anyone jump on a dot-0 release of anything. But Apple aggressively attacks dot-0 bugs — iOS 11 is already at 11.1.2[1], and I'm running an 11.2 dev release that's been nicely boring for me.
Apple has had a planned-obsolescence approach for years and years [1], but with the way software used to be distributed and the consumer expectations on hardware design it was less of a problem before. If you want to keep using an old Mac, there are plenty of OS install disks, copies of old shareware floating around, sites like Low End Mac which document what works and what doesn't. The hardware was also usually more repairable. I'm worried that current Macs in the age of App Store OS updates and glued battery will be rendered into expensive paperweights that much more certainly when Apple withdraws support.
You may laugh, but it seems scarily plausible to develop a machine algorithm that generates 12 articles, have a human go through to pick the 3 least-ridiculous ones and do some light editing, and push it out to the news.
Motley Fool seems to be the leaders in this -- if you search google news for any stock you're interested in then you'll find that Motley Fool seems to be cranking out largely-boilerplate articles about it; sometimes they seem to have had some human intervention, other times they seem to be fully computer-generated.
I was seeing credible copy produced for financial reporting in the late 1990s.
The underlying data are available, the vocabulary is distinctly limited, and the reporting was already highly pro-forma. Mad-libbing a few variants of language and selecting the best of the lot does work fairly well.
That actually raises the underlying question: why publish narrative copy at all in cases such as this, rather than data tables or charts? News and journalism are curiously allergic to data or presenting it in a usable fashion -- I've seen tables essentially written out over several paragraphs that could have been expressed in a few grid squares.
Routine financial news, sports... Anything that can be largely churned out in boilerplate fashion based on some standard facts. This has been done for a while now and it will only increase as long as there's a business model for quickly published, largely undifferentiated, rote stories. It's a low-cost, high-volume game.
