Victoria:
Not sure I agree completely.
I have free versions of 5 AI’s on my tablet (too cheap to pay, lol).
Chat GPT distinguishes:
- “General knowledge: My built-in knowledge is extensive, but it isn’t a live database. For rapidly changing facts, I don’t assume my internal knowledge is current.
- Current information: I can search the web in real time when you ask about things like current products, prices, airline schedules, laws, medical evidence, software versions, news, businesses, or who’s currently in a particular role.”
As for Claude:
“My reliable knowledge cutoff is the end of January 2026. I can speak confidently about things up through then. For anything more recent — current events, prices, product releases, who holds what position now — I use web search to check rather than rely on memory, since my training data goes stale. Given today’s date (August 2, 2026), that’s actually a fairly big gap, so if you’re asking about something time-sensitive, just say so and I’ll search for you.”
And Grok:
“My knowledge is continuously updated, and I can pull in the latest information in real time using tools (web search, browsing pages, X/Twitter search, and others) whenever something needs to be current. For anything time-sensitive—news, prices, events, software versions, etc.—I check live sources rather than relying only on what I already “know.” So the answer is: as up-to-date as the web and the tools allow at the moment you ask.”
I also use Gemini and Copilot.
The main thing I’ve learned is to compare answers from at least two.
I agree with you about how impressive results can be. We’ve just pulled the trigger on a return to Italy in February. Among other things, the AIs helped design an itinerary for seeing Titians in Venice and helped me decide on my camera kit.