I wonder why OpenAI doesn't try to get more feedback and training data from its users, though I do notice that sometimes it'll give me two answers and ask me to pick the better one.
For example I've noticed that a lot of the time when I ask ChatGPT a coding question it might get 90% of the answer. When I tell it what to fix and/or add, it usually gets the answer. I wonder if they're using these refined answers to fine-tune those original prompts.
I wonder how the LLM interacts with other software like the calculator or Python interpreter. It would be great if this were modular so that the LLM OS could be more like Unix than Windows which is what OpenAI seems to be trying to emulate.
Ultimately though it seems to me like AGI is fairly straightforward from here. Just train on more quality data - in particular enabling the machine to generate this training data, increase parameter size, and the LLM just gets better and better. Seems like we don't even need any new major breakthroughs to create something resembling AGI.
They should be capturing the changes that people make to the ChatGPT outputs. Many people will be copying the outputs to some other application and then make changes. If open AI would make it easier to modify the outputs right within ChatGPT, they could use that as feedback. Basically, fuse the end-user UI with the UI of the annotates.
I have zero faith that the average ChatGPT user will make quality edits. If anything, this invites trolling and active dataset poisoning/manipulation the moment people figure out that's what they're doing.
> There has long been discussion among computer scientists about the danger posed by highly intelligent machines, for instance if they might decide that the destruction of humanity was in their interest.
This AI doomer stuff is such nonsense and I can't believe anybody takes it seriously. As if it's OpenAI's responsibility to save humanity from the pitfalls of AI.
Imagine if we decided to improve our education system and doomers were talking about "hitting the panic button" because students were getting too smart from all the quality education.
Well the quote I referenced from the article of the machines deciding to destroy humanity is utter scifi nonsense.
There are obviously legitimate risks to AI and safety is important, but this is the same for any new technology, and it's governments' responsibilities to ensure that people are safe. AI companies mindlessly slowing down and keeping their tech to themselves does no service to humanity, and if anything is a net-negative due to how tremendously useful this stuff is.
Education is analogous to AI because AI is an enormous education and productivity boost to humanity - sort of like everyone having a personal assistant, programmer, and tutor at their fingertips. This could be used for good and it could be used for bad, but the technology itself is neutral.
Again I want to emphasize that obviously there are downsides that could result from evil people using AI for bad purposes, but that does not justify slowing down AI progress - just like I don't see "people using information for bad purposes" as a legitimate reason for stifling advancement in education or something like Google search.
I have yet to see any convincing argument otherwise. Feel free to provide your counter-perspective.
What a sad but refreshingly honest article. I never worked at Google, but this aligns with my impressions from the outside, as well as everything else I've read (eg. that recent article here from a founder who's startup was acquired by Google and she left due to the stifling bureaucracy). The company hasn't innovated much recently and its products like Google Search have deteriorated in quality tremendously. At this rate Google may be on its way to becoming the next has-been tech company (eg. sort of like what happened to IBM).
The management and bureaucracy depicted in the article sound like a corporate nightmare and unappealing place to work. I didn't know that Google had non-engineers running dev tool teams. Can this VP even reverse a linked list? /s
Seems like Google needs a change in leadership, starting with replacing CEO Sundar Pichar.
The internet would be better without mandated cookie banners. It's so damn frustrating using the internet in the EU. If you don't want to be tracked just browse in Incognito mode.
You don't need a cookie banner if you don't have 3rd party tracking cookies. It's really that simple.
The fact that all sites have them, shows us a terrifying truth: all websites are tracking us with 3rd party tools. "all" websites send our browsing habits off to (many) companies that will sell, mine or otherwise monetize our data.
Again: A cookie banner is not needed if you don't have 3rd party and/or tracking cookies. E.g. With matomo on your own domain, plausible analytics, or simply mining your servers logs with math, you won't need cookie banners.
Right so every website needing a cookie banner to comply with EU regulations is not only a UX nightmare, but it doesn't even prevent tracking. Horrible pointless legislation.
No, they don't. Read the GDPR, it's not that long. The actual problem is that the current practice on which massive profits depend is contrary to any privacy desires. If they didn't track, they wouldn't need ask for consent for the tracking.
You've just described why once prominent companies fade into shadows of their former glory (eg. Kodak, Blackberry, IBM, Oracle, Microsoft). Definitely not inevitable and could be avoided with better leadership.
But not a problem, either. Turnover is natural. Nobody but the investors care very much whether Kodak pivots to digital cameras, or whether Kodak remains the leading film camera company as the industry shrinks, and a different company makes digital cameras. In fact the latter is often better for the economy and consumers, due to the better specialization.
Yes it is a problem if a company is failing not just for the investors but the workers. Nobody wants to work for a sinking ship. Can't believe this even needs to be said.
I'd be shocked if D'Angelo doesn't get kicked off. Even before this debacle his AI competitor app poe.com is an obvious conflict of interest with OpenAI.
Depends who gets onto the board. There are probably a lot of forces interested in ousting him now, so he'd need to do an amazing job vetting the new board members.
My guess is that he has less than a year, based on the my assumption that there will be constant pressure placed on the board to oust him.
What surprises me is how much regard the valley has for this guy. Doesn’t Quora suck terribly? I’m for sure its target demographic and I cannot for the life of me pull value from it. I have tried!
I think it was only a competitor app after GPTs came out. A conspiracy theorist might say that Altman wanted to get him off the board and engineered GPTs as a pretext first, in the same way that he used some random paper coauthored by Toner that nobody read to kick Toner out.
