Beep boop - this is a robot. A new show has been posted to TWiT…
What are your thoughts about today’s show? We’d love to hear from you!
So many good topics today.
AI Regulation: I think we are already seeing one of the problems with the lack of regulation, or rather “Big Tech”, in general’s disregard for existing laws. “Those are laws for the little people, they don’t affect us!”
This has little to do with AI specifically, but the general attitude of big business, and especially Silicon Valley Start-up Culture. This problem goes back a couple of decades already. We have companies like YouTube, and through their purchase Google, who blatantly ignore laws and the law makers turn a blind eye, because they are a start-up, so “give them some leeway, they need time to grow.” But that leeway is never clamped down upon, until it is too late. The same goes for Meta etc.
They start by ignoring laws, such as copyright, we need growth, we need to get content out there. People say, “but you are showing copyrighted material,” they say, “so what? If we had to worry about that we couldn’t grow!” So they get a “temporary” pass, which becomes institutionalised, until they are so big that it is a problem, but then they whine, “but we have so much content, it would be impossible to police it properly,” (i.e. it would cost us most of our income to sort this out properly and compensate people correctly). So they get away with not doing the job properly, they end up paying a pittance for their content and the reporting systems are broken by design, so the people with a genuine beef have problems getting compensated for all the content that is stolen from them and on the other hand, serial offenders can make bogus claims and get genuine material incorrectly pulled down (I know that TWiT suffered this for a long time on YouTube).
Move forward to the AI start-ups and we (the governments, law makers and general populace) have learnt nothing from the complete catastrophe that was not reigning in big business when they were small and there was a chance to force them to do things properly and have the processes in place before they grew, as opposed to waiting for them to grow, then listening to how it is impossible to implement any checks and balances at scale. And we are doing the same thing with AI, the LLMs are openly gobbling up illegal content (copyrighted works of all forms) without first getting licenses and paying compensation to the creators and their agents (yes, that system is broken too).
They get a pass, “but the content is on the Internet,” well, yes it is. But a lot of it is on illegal sites or has been posted to sites like YouTube illegally and they have failed to pull it down, for example. Two wrongs don’t make a right. They have knowingly used illegal content in their training, but they don’t care and authorities are too scared about “squashing innovation” to actually do anything to reign them in - yet now is the time when we can do this, before they are too big and too institutionalised. The LLM business is still nascent, it is less than a year old (commercially, with major AI products). Forcing them to put filters in now, so that known sites with illegal content or on content that hasn’t been licensed aren’t used in the model, for example, and to re-train their models on the “correct” dataset. If we wait to see where the industry goes, it will be too late, again!
We can all see the problem, but instead of finding a solution, it is, “don’t squash innovation, we’ll sort it out later, if they are a success.” But history has shown, time and again, that the wait-and-see approach has just made the situation worse, the law breakers have earned so much money that they can afford to lobby that they shouldn’t be prosecuted for breaking the law, in fact, the laws should be weakened to suit them!
Chargers I don’t understand Mike’s problem. When I used to travel, I used to have a private Nokia phone and a company Motorola, I’d have my Olympus camera and my Canon Camera, a laptop, torches and various other electronic devices. Each of those had a different, unique connector for charging, each would be directly wired into the charger, so no only would I carry the devices, I’d carry a dozen different charges and cables to connect things up.
Now, I just need one charger and one cable for everything (well, I still need to replace my Sony camera, which is micro-USB and I still have an iPhone and Apple Watch, so I actually still need 3 cables, but still just one charger). It is the same in the house. I have thrown out around 20 different chargers and cables over the last 12 months, I am down to a single charger in my office in the basement, a charger in the kitchen and a lamp in the bedroom that has a USB-C port for charging other devices over night. I don’t have to carry cables and chargers from room to room because they are all different.
The EU tried to get companies to do this voluntarily, they gave them nearly a decade, before they stepped in and put it into law, because industry had shown they couldn’t do it themselves. And the argument that “USB-C might be better than Lightning now, but what about the future?” The law in question has a clause that the state of connection and charging will be regularly reviewed and if better standards come along the law can be amended to take that into account (E.g. all devices from 2022 - 2035 have to be USB-C, all devices after 2035 have to be TRA-B - yes, that is a deliberate sci-fi joke).
1 Like
I also signed into Facebook for the first time in several months to view an event invite for a religious conference. The news feed is unrecognizeable, between the advertisements and things “you may be interested in”. It bears little relation to the product I signed up for 15 years ago.
Apparently because I viewed a conference that was out of town, that gave the algorithm a signal that I wanted everything to do with that city. Concerts, yard sales, churches, whatever.
I’ll stick with Twitter/X tbh. At least they show me posts I subscribed to, some ads and little else.