Businesses do not have an entitlement to profit. Suspending customers for using a fairly expensive subscription plan -- especially forfeiting an annual prepayment for a day or two of coloring outside the lines -- sure does make Google appear entitled to profit without ever risking its own pricing model.
> Suspending customers for using a fairly expensive subscription plan -- especially forfeiting an annual prepayment for a day or two of coloring outside the lines
they're being suspended for using a private api outside of the app for which the api was intended. If you make a clone of the hbo app, so that you can watch hbo shows without ads by logging in with your discounted ads-included membership, your account will also be suspended.
The facts are straightforward, even without analogies. But since we're using them...
You are at the grocery store, checking out. The total comes to $250. You pay, but then remember you had a coupon. You present it to the cashier, who calls the manager over. The manager informed you that you've attempted to use an expired coupon, which is a violation of Paragraph 53 subsection d of their Terms of Service. They keep your groceries and your $250, and they ban you from the store.
Google is acting here like it was entitled to a profitable transaction, and is even entitled to punish anyone who tries to make it a losing transaction. But they're not the police. No crime was committed.
Regular businesses win some and lose some. A store buys widgets for $10 and hopes to sell them for $20, but sometimes they miscalculate and have to unload them for $5. Overall they hope their winners exceed their losers. That's business.
my point wasn't an analogy. the facts are that it is a private api being used with a subscription service. neither hbo nor google are required to do business with people that abuse the api.
We are in violent agreement about that point. Where we seem to disagree is that I don't think they're entitled to also keep the customer's annual subscription payment when they've decided they want out of the contract.
Equally, customers are not entitled to make set the terms, or pricing decisions for businesses. They can always move their custom elsewhere if they disagree with ToS or pricing.
The best take I've seen on the whole `AI will replace all devs' is a way for big tech to walk back the disastrous over hiring they did around Covid without getting slaughtered in the stock market.
However, admitting to have massively over hired and wasted a lot of money does not make the management involved look good. No one wants to admit they made a massive blunder.
The market doubly rewards companies that lay off workers and have a story about how they're automating everything with AI, even if that story is just a story.
I don't know who the author talks to, but everyone I know who does anything with gen Ai visually uses comfyui: https://www.comfy.org/
The end result is that you're hugely more productive, but need to be just as skilled as you were in the old days when you were working with photphot or after effects.
My funniest moment working in Singapore was translating between an Indian and a Chinese co-worker. The translation was repeating what each said in English in English.
I use a screen reader and in managed quite well until 1200.
That said: phonetic spelling now. We have spent 500 years turning English into something closer to Egyptian hieroglyphs than a language with an alphabet.
You suffer from the typical brain damage caused by using a language without an alphabet.
There is no such thing as spelling in phonetic writing systems because they render what is said, not some random collection of glyphs that approximated how a word was pronounced 500 years ago, in the best case.
If two people with different accents can speak to each other, they can also write to each other under a phonetic writing system.
Then under your definition there must not be any widely used written language with an alphabet. Most of the world's alphabetic writing systems aren't phonetic transcriptions, they're standardized. They're usually based on the prestige dialect, at the cost of diminutizing other dialects.
For example, Spanish has a fairly consistent spelling system standardized by RAE, based in Madrid. But, for instance, even though much of Latin America doesn't have a distinction between s and soft c (seseo), they still keep the distinction in its spelling.
One I can say for sure is Serbian. Italian looks like it does. Finnish, Hungarian, Georgian, Armenian, Albanian, Turkish and Korean are all ones I've heard are to a lesser or greater degree, but I don't know enough to say either way.
People always overestimate how 'phonetic' their language is, because nobody actually uses phonemes in regular speech. In Korean in particular, there doesn't even seem to be any obvious correspondence between what is written, and what is actually said.
Foreign accents don't come from any inherent inability to learn language after X years of age. They come from people pronouncing languages as they are written, and virtually no language is like that in reality.
It's true that when studying a foreign language, learning to read too early can harm your pronunciation. However, it is very difficult to learn new sounds that have no equivalent in your native language, and some languages have very restrictive phonology (like Italian and Japanese requiring a vowel at the end of every word) that their native speakers struggle to break out of.
Standard Italian speakers in Rome struggle to understand Ciociaro dialect, which is from the region on the outskirts of Rome. Take "n'coppa" - spelled with a "c" but very much pronounced /ŋgopa/ with a voiced [g]. I dont even have a reference point for Sicilian but that really pushes the bounds of the dialect/language distinction.
That's one example, from a language with ~70M native speakers, in a geographically tight region.
Likewise, all your other languages (sans Turkiye) are very compact geographically with small speaker bases. And Turkish undoubtedly has large aspects of forced standardization and dialect extinction.
English is spoken by 1.5 billion, by ESL speakers from basically every language tree, across the world. Try to get folks from Boston, Brooklyn, Philly, and Albany in a room and get them to agree on a phonetic spelling.
That's kind of a mean and not very relevant response.
The point is that if anyone wanted to reform English spelling, they would have to choose a particular dialect to standardize around.
There is no standard English dialect. There is a relatively standard version of American English ("Walter Cronkite English"), and there is Received Pronunciation in England, but then there are all sorts of other dialects that are dominant elsewhere (Scotland, Ireland, India, etc.).
Which one should we choose to base our orthography on? Or should we allow English spelling to splinter into several completely different systems? Yes, there are already slight differences in British vs. American spelling, but they're extremely minor compared to the differences in pronunciation.
And after this spelling reform, will people still be able to read anything written before the reform, or will that become a specialized ability that most people don't learn?
Worked for thousands of years with other phonetic written languages. Words change spelling over time, instead of pronounciation drifting without the spelling changing.
You're proposing to make reading just as difficult as understanding every other dialect of spoken English - something even most native speakers have difficulty with.
Your proposal would also eliminate whole-word recognition, which is what makes reading fast. It would slow us all down to the speed of young children just learning to sound out the letters.
Right. Because everyone gets confused when you write behavior instead of behaviour or license/licence or analyze/analyse. It’s so confusing that there are already different ways to spell the same thing.
American English isn’t the only spelling of English.
There are exactly two ways to write license. What you're proposing is that there should be 20 different ways to write it, depending on what particular dialect of English you speak.
We don't really manage it with speaking. I don't understand highland Scottish dialects at all. I have trouble understanding Cockney.
Yet people who speak those dialects can write anything down and I'll understand it perfectly with no effort.
You don't understand the value of standardization. It's what makes reading fast and independent of dialect. People who read English don't literally sound out the letters. They recognize the whole word instantly. Sounding out the letters is only a fallback mechanism.
What you're proposing might work for a tiny language with only one main dialect. English is a global language with a huge number of dialects. Major languages like this need standardized writing systems, and to no one's surprise, they all have them.
> Around 1809, ... Sequoyah began work to create a writing system for the Cherokee language. ... He worked on the syllabary for twelve years before completion and dropped or modified most of the characters he originally created.
> After the syllabary was completed in the early 1820s, it achieved almost instantaneous popularity and spread rapidly throughout Cherokee society.[4] By 1825, the majority of Cherokees could read and write in their newly developed orthography. ...
> Albert Gallatin ... believed [the syllabary] was superior to the English alphabet in that literacy could be easily achieved for Cherokee at a time when only one-third of English-speaking people achieved the same goal.[6] He recognized that even though the Cherokee student must learn 85 characters instead of 26 for English, the Cherokee could read immediately after learning all the symbols. The Cherokee student could accomplish in a few weeks what students of English writing might require two years to achieve.
Phonetic spelling would perhaps make the language easier to learn for native speakers, but it would make it harder to learn for foreigners, at least those of us who come from Europe. Most words in written English resemble words in Germanic or Romance languages. If English was spelled phonetically, the resemblance would be significantly smaller.
People often say that the English spelling is weird or illogical. As a non-native speaker, I disagree. The English spelling makes perfect sense. It’s the English pronunciation which is really strange and inconsistent.
> Phonetic spelling would perhaps make the language easier to learn for native speakers, but it would make it harder to learn for foreigners, at least those of us who come from Europe.
BS. Phonetic alphabets are _much_ easier to learn for everyone. In Russia and Ukraine pretty much every child can read by the time they enter the first grade. It's _that_ easy because both alphabets are phonetic (although it's only one-way in case of Russian).
Meanwhile, when I was learning English there basically was one spelling rule: memorize. It was not at all helpful. I also ended up learning English as a mostly written language, so after moving to the US, I kept getting surprised by how familiar written words are actually pronounced.
E.g. it took me a while to explain to a nurse over the phone that I may have pneumonia and need an appointment. Why the heck that leading "p" is completely silent?!?
> In Russia and Ukraine pretty much every child can read by the time they enter the first grade.
In the US too, reading is generally handled in Kindergarten, the year before first grade. If your parents didn't teach you before that, like mine did.
> Meanwhile, when I was learning English there basically was one spelling rule: memorize.
There are rules though, that we're ad-hoc taught as kids, or just absorb through exposure. Just because there's a lot of exceptions doesn't mean they don't exist. Here's an attempt at listing them out: https://www.zompist.com/spell.html
To be technical: the term is phonemic, not phonetic. If we spelled phonetically, we'd have different symbols for the p in 'spin' and the p+h in 'pin'. Similarly for 'tick' and 'stick', and 'scale' and 'kale'. Native English speakers generally don't notice the differences, just like speakers of many oriental languages don't easily recognize the difference between English /l/ and /r/.
The usual pictures of и / п / т / ш ambiguity that you see are exaggerated in that they show forms that are nominally “standard” but basically impossible to reproduce without a fountain (or, even better, dip) pen (think round hand or, as 'cyberax mentions, Spencerian script), yet use a constant stroke width that such an implement wouldn’t produce. For the latter two, people who actually write m and not т will often resolve the ambiguity with ш with an over- resp. underbar (the same ones that Serbian uses even in print[1]). It’s also pretty normal to exaggerate letter joins when they come out looking too similar to parts of other letters, etc. Overall, modern Russian cursive is about as legible as the modern French one, and I don’t think people complain much about the latter.
I also find the hand-wringing about English accents somewhat surprising. Yes, different accents exist, and yes, English has a much wider variation than (urban) Russian (there are things in the countryside that urban dwellers haven’t heard for a century), but phonemic orthographies are a thing, and though children in e.g. Moscow may perpetually struggle with orthographic distinctions that no longer correspond to anything in their accent, the idea of a spelling competition remains about as laughable as that of a shoelace-tying one. Nobody makes you represent the many mergers of English with a single letter in your new orthography (though it would be funny).
It depends on a writer, but it can be very legible.
I used to be able to jot down notes during lectures almost as fast as the normal spoken speed. We often traded notebooks when preparing for the exams, and I rarely had problems reading other people's notes.
It's also really nice to write, once you learn it. I was surprised after moving to the US that almost nobody here knows how to write in cursive anymore.
I've been thinking about this problem for quite awhile, and recently coded up something that allows for easy conversion between today's written English, and a phonetic spelling convention.
I think he means the latter. This makes learning the spelling harder because you have to learn each word individually, as you would have with hieroglyphs, as opposed spelling it out based on phonemes (that you would have learned from learning how words sound when spoken) and a limited alphabet.
That's not how I learned to read or spell in the 1970s. "Sounding it out" was the main strategy. You learned a few rules for how different combinations of letters sounded, and the exceptions to those, as you went along. But most words are spelled as they sound.
That's why kids start with "Run Spot Run" and other simple 3 and 4 letter words. They then learn the more complicated rules and exceptions as they go. It's really not a problem.
Spelling can still be phonetic even if groups of letters have differing sounds from those letters' sounds serially in isolation. The key criterion is that the rules must be universal, applying to every instance of those groupings, rather than having exceptions for their appearances in certain words.
...ok, it occurs to me now that a smart-alec might declare each individual word to be a "grouping of letters with its own phonetic pronounciation", whereupon phoneticism as-defined is achieved trivially because pronounciation is universal over the singleton universe of words spelt exactly like that word. You know what I _mean_ - "sufficiently small groups of letters", hand-wave.
The issue is that the language can never render that collection of letters. Sh in English can render the sh in sheep. It can't render any word with the sounds s and h together.
An alphabet assigns a letter to a sound. No more no less. English no longer has an alphabet because the Latin alphabet, designed for Latin languages, replaced the native Runic alphabet.
Serbian has an alphabet, as does Italian. All other European languages I'm aware of don't.
I'm guessing the first question will be "How are we going to keep the UI consistent?". The hard part is never the code writing it's carefully releasing fast changing features from product people. Their chat UX is the core product which is replicated on the internet and other devices. That's almost always React or [JS framework] these days.
Migrating the system would be the easier part in that regard, but they'll still need a JS UI unless they develop multiple teams to spearhead various native GUIs (which is always an option).
Almost every AI chat framework/SDK I've seen is some React or JS stuff. Or even agent stuff like llamaindex.ts. I have a feeling AI is going to reinforce React more than ever.
Yep, I understand why let's release this one feature everywhere is a great lure and I do get annoyed when desktop vs mobile spotify gets features later or never. However, a phone is not a desktop capability wise and what we usually get is the power of the phone on a desktop, aka the lowest common denominator of capabilities.
This fetish we as an industry have to hide platform specifics makes us blind to the platform specific capabilities. Some software would be better off if it leaned into the differences instead of fighting them.
The corollary to the bitter lesson is that in any market meaningful time scale a human crafted solution will outperform one which relies on compute and data. It's only on time scales over 5 years that your bespoke solution will be over taken. By which point you can hand craft a new system which uses the brute force model as part of it.
Repeat ad-nauseam.
I wish the people who quote the blog post actually read it.
>In software San Francisco is still the top for AI research
What was the last thing that a major US Lab published? It's all trade secrets.
Chinese labs are the only ones publishing results as they happen.
The US is in the position it was for semiconductor manufacturing, first it was labs and open science. Then by the 80s fabs started costing millions and universities stopped being able to contribute and nothing got published.
Now it's getting to trillions and if Intel goes under there is no one in the US who knows how to make any semiconductor generation newer than 2010.
llama hasn’t had a new version in over a year. off the top of my head there are at least 4 entire new series of Chinese based llms that have been open sourced
Yet another lie by ommision. Violent deaths by guns have no relation to strength of gun laws. What your link measures is the number of accidental deaths by guns. If gun owners want to kill themselves it's not my job to keep them safe.
> Suicidal people aren't a valid reason for my rights to be restricted, sorry.
You also have a right to travel around the country, but that doesn't mean you're allowed to drink and drive. There are plenty of valid, constitutional reasons for firearm ownership to be restricted to qualified individuals. When these restrictions are in place, many fewer people die. It is what it is.
According to the first militia act, every able bodied male over 18 is what defines a qualified individual. Beyond that, you're actually required to own a firearm in that case.
So are you advocating to outlaw alcohol? I mean, since people get depressed and drink which drives more depression and kill themselves... I guess you're suggesting that all depressants should be outlawed.
What the hell do you expect? To get paid for using other people's tools on Google's servers?
reply