GPT-4 is out | A good old fashioned bank-run! | Online Safety Bill VS WhatsApp
Sorry I didn’t send a newsletter last week, I was too busy uh… not writing it. This issue is pretty long though so, kind of get over it.
This week was I dunno, fine?? Could Have been better? 🤷♀️. Because:
GPT-4 has been released; we don’t know how it works or if it’s even useful; watch as it eats the world
Silicon Valley Bank has collapsed; watch as tech bros scramble
WhatsApp and Signal will not comply with the Online Safety Bill; watch as these two products leave the UK completely!
😘 GPT-4eva xoxo
GPT-4 has just been released, so naturally my Twitter feed (full of ‘popular’ accounts that I do not follow) is now spitting out threads from white guys showcasing what ‘incredible things’ other white guys are doing with it already. Those ‘incredible things’ include ‘one-click lawsuits’, drawing a wireframe and turning it into a functioning website, and finding problems in an Ethereum smart contract. So basically: all things that are connected to making GPT-4’s very privileged users even richer.
I have to fit in writing this newsletter around other work, so I don’t have all the time in the world to form thoughts and opinions about the gushing torrent of ‘innovations’ coming out of generative AI right now — it’s all happening pretty fast (even the phrase ‘generative AI’ is not one I was even using less than six months ago). But the top-level concern that seems to always come back to me is this: once this god-awful fever dream of hype finally dies down, we will be left with another ‘normal and every-day’ piece of technology that perpetuates neo-colonialist practices right before our eyes.
Honestly, having my online experiences ingested by a machine without my knowledge only to have a bunch of tech CEOs blithely sell the outputs back to me in the name of technological progress is a very 2023 version of gaslighting. It’s weird how we’ve spent the last couple of decades on this ‘new’ thing called the internet, interacting with each other, creating new and amazing things, and having a great time, and now we’re being told ‘don’t worry — there’s a machine that can do that for you’.
Every other tweet I’m seeing this week is either talking about the fact that GPT-4 can pass the bar exam, or that the 98-page technical document about it manages to completely avoid explaining how it works. Two things from me:
Literally who cares that it can pass the bar exam. What are we as a society supposed to do with that information? Stop training to be lawyers? Laugh and point at the lawyers we already have? It’s a technological marvel! It can pass the lawyer exam but it can’t actually be a lawyer!
If I ever meet Sam Altman I’m gonna ask him to change his company name to ‘ClosedAI’ and also to fuck off forever. I think that with OpenAI’s products, there has been a conflation between ‘everyone can try this for free’ and ‘open source’ — and this is actually a huge problem.
‘Open source’ is not where you just let people use your gruesome product for free, even though you have no idea what effect it will have on society. It’s where people can actually look at the source code of what you’ve made, understand how it works, and potentially make it work for them. That is not what’s happening here. This is a completely walled off piece of technology with a paid premium tier.
The generative AI hype has also actively made existing digital products and services worse. Literally every single tiny thing I use has now been injected with pointless show-boaty AI integrations. Every single context menu in Notion (the thing I use to write this newsletter) tries to get me to improve my writing with the use of it’s in-built AI. These context menus are literally just lists of things that clients pay me to be good at (e.g. ‘summarise’, ‘simplify’, ‘change tone’). Weirdly, I am NOT into it, not one bit.
Hey, it generally takes me ages to put Horrific/Terrific together every week; I love doing this but I would love it even more if you could chuck me a wee bit of money every month — thanks you’re such a champ :)
🤡 Maybe all of you shouldn’t have used one bank for everything, just a thought
As you may have heard, the preferred bank of choice for tech entrepreneurs, Silicon Valley Bank, has now shut down. This was a bank that startups used to store all their VC money, and I guess also use for revenues and overheads, oh and also payroll. It’s funny how the people on this Earth who are the absolute loudest about being different and innovative decided to flock to the same bank only to make it collapse under the weight of their short-sighted greed.
Other people have followed the story closer than I have, but essentially:
VC’s and entrepreneurs failed to be responsible with their millions of dollars, and literally kept all their money in one bank (SVB)
SVB invested these funds into government bonds, which recently lost a hell of a lot of value because of [boring stupid financial problem that shouldn’t exist]
Naturally, everyone got on Twitter and WhatsApp and started panicking about their money, catalysing a run on the bank.
I actually don’t understand how you can be this negligent with your money, especially if money is the single most important thing to you. These accounts were only insured up to $250k — and over 93% of deposits in SVB were significantly bigger than this.
Furthermore, it’s very interesting to see this horrific rash of men completely abandon their alleged neo-liberal or libertarian ideologies, and start demanding that the president guarantees the safety of their money — as if the fate of the world depends on it. Next time that you think anyone in Silicon Valley believes in something, even their purported ideology, remember this week of shameless hypocrisy. Men who ‘make their own luck’ apparently do so by asking the government to bail them out. It’s not a ‘hand-out’ if you’re already drowning in money!
🙅♀️ WhatsApp and Signal would rather leave the UK than comply with online safety bill
The UK’s flagship internet regulation, the Online Safety Bill, will apparently give Ofcom powers to look at your private messages and check that none of them include child sexual abuse material (CSAM). Yes yes please invade my privacy… as long as it’s for the children.
🤦♀️ Very obvious fact that has alluded the UK government: in order to peek into private conversations, you’d have to do away with end-to-end encryption.
There’s a reason why this bill has been re-written and shat on a million times over by the endless stream of inept cabinet ministers have slithered their way through parliament: literally no one working on it knows what they are doing. A private messaging service like WhatsApp or Signal or iMessage can’t just ‘remove’ encryption for one country — it just doesn’t work like that. If messaging services were to actually follow this rule, they would have to drop encryption altogether. For everyone everywhere. It’s a completely unreasonable demand.
Both the heads of WhatsApp and Signal came out last week saying that they aren’t going to significantly weaken the privacy and security of their platforms to suit the UK alone. So, if this rule remains and the bill actually gets passed into law, both Signal and WhatsApp will just have to stop operating in the UK to avoid facing the legal ramifications. If you consider that 98% of WhatsApp user’s are outside of the UK, it’s very strange that the UK government think that their country would be a priority at all — literally, why would an internet-based service bend over backward to comply with this law?
Now, I’m pretty sure the government’s intention here was not to inadvertently delete two key communication services from the entire country, but I’m also pretty sure that they all have Eton Mess for brains. So… let’s just see what happens.
🍨 Had enough? Too bad, here’s some DESSERT
Here are a few other things that caught my eye this/last week.
Because they never have any of their own ideas anymore, Meta are going to a build a ‘federated’ social network that is text-based only. Kind of like Twitter. And kind of like Jack Dorsey’s federated alternative Bluesky, which I wrote about a couple of weeks ago.
India are ramping their misguided paranoia up to a million by proposing that smart phones are no longer shipped with pre-installed apps, and that the Indian government should be allowed to screen any major operating system updates. They say that native apps pose a security risk, but really this just seems like a reason to intrude on the personal devices of their citizens.
The Biden admin have finally realised that their heavy reliance on just a handful of cloud infrastructure providers is a security risk, and probably some regulators should get involved. Can’t wait to see what janky new laws come out of this!
Thank you for reading, and remember to keep you and your family safe by permanently banning TikTok from your household. Goodbye.