OpenAI Finally Allows ChatGPT Complete Internet Access
gizmodo.com
gizmodo.com
My problem with it is that it seems to run really obvious, naive searches. Most of the time I ask it something, then see what it's searching for and think "oh no, that's not going to return anything more useful than what I could have found myself".
It's also pretty slow.
I'm very much looking forward to having a search assistant which can go ahead and wade through eg 20 websites about home battery packs and show me a summary table of my options - ChatGPT Browse isn't quite that yet.
Analyzing and interpreting new information is harder than regurgitating what one already knows, so perhaps this isn’t surprising.
Even gpt-4, as amazing as it is, has quite limited analyzing, reasoning, and interpretation capabilities, compared to skilled humans.
Of course, the fact that we have to even ask this question is kind of unbelievable…
Chatgpt4 seems most powerful but Claude offers unlimited context which can be used to summarize entire books
Prepexity can browse and summarize web
Which one should I subscribe to?
I just tried it out with my normal test queries and found it to be far worse than Google's Bard.
I asked it to tell me when the next baseball home game is so I can avoid traffic when going to the downtown library and it couldn’t answer even that basic question. The search results defaulted to the team’s official calendar but the day numbers were displayed as images so they could style them in the team’s font, making them invisible to ChatGPT.
Relatedly, in my experience, for some reason most internet properties relating to anything sports related (MLB, NFL, NHL, or even small town little league) have absolutely terrible, extremely over complicated user interfaces.
Kagis' FastGPT is pretty good in that regard (it uses kagi search + claude from anthropic) and it's really fast. Their other AI search/chat which is in beta currently (for ultimate users) is even better IMO.
1. I don't want to read through all materials in that topic but want an overview.
2. I can't describe the topic with the precise language for it, hoping ChatGPT can translate my naive description to the vocabulary for that field.
ChatGPT does bad in both of these cases. For (1) it just randomly browses 3~4 search results, as for (2) it always try to search in my original vocabulary. Why don't I search myself then, but to wait for the slow response?
A ninja battle in the style of Bob Ross is going to be some trees and mountains while the ninja battle is assumed to be happening out of view on the other side of it.
This started with scrapers and aggregation sites and has gotten noticeably worse.
https://platform.openai.com/docs/plugins/bot
User-agent: ChatGPT-User
Disallow: /Not really. That use case is done, of course, but the primary purpose of robots.txt is to help crawlers by indicating what parts of the website are appropriate to be searched and what parts aren't.
Robots.txt is not intended primarily as a means to defend a site against crawlers. That's why it relies on the goodwill of crawlers to work.
There are two agents, but apparently you can't allow one but disallow another.
For example if you make a web page a user pulls up that calls another webpage, is that a user action, a script action, a mix of both? I personally would consider it a user action.
Challenge accepted!
There's an option in Bard's settings to enable real time replies.
Just got access to Bard API and hoping that continues delivering what I have been happy with...
1. Scam people over the internet to get money
2. Hire hitmen to kill its makers.
3. ?
4. Achieve world domination.
Default (GPT-3.5)
User do you have access to search the internet yet? ChatGPT I do not have the capability to search the internet or access real-time information. My knowledge is based on the text that I was trained on, and my training only includes information up until September 2021. I can provide information and answer questions to the best of my knowledge up to that date, but I cannot browse the web or access current information.
Its training data predates its creation, and the model doesn't get updated every time they ship a new feature for it.
See also: https://simonwillison.net/2023/Mar/22/dont-trust-ai-to-talk-...
3.5 can't browse, only 4 so the above is perfectly correct.
If you were repeatedly flash frozen and flash unthawed to be asked questions about yourself, with the same memories upon thawing as when frozen, for those moments you weren't frozen would you be "remotely conscious"?
I think you'd think you think, don't you think?
I think you’d be able to answer the question, don’t you?
I suspect that any case of them not honoring it is probably due to your site content being available on archive.org or common crawl or some other service.
It's a lot like saying, "if it's on Github, I can ignore your LICENSE file" or "if it's on DA I can ignore your CC-NC license." Both of these seem technically enforceable (if not expensive) so I'm not quite understanding what OpenAI is grounding their justification in.
It seems much more likely that OpenAI stands to make a lot of money by arguing that it's fair use and then back filling in an argument rather than making one from first principles.
OpenAI leads the cult.
This is why I took down some of my sites, and put a login in front of the rest. Until I have some solid means of defense, I can't think of any other effective approach.