OpenSource will always win over closed source, so OpenAI (AS THE NAME SAYS) used to be (at least partly) OpenSource available accessible for every user (with the right hardware)

it is not true that deepseek can be run on a nokia flip phone, it needs massive computational resources and then it’s still slow X-D and this AMD EPYC 7402P @ 2.80 GHz 1 Processor, 24 Cores 48GB RAM is hitting its limits
it works 😀 but it ain’t fast: a few tokens per second:
“2023, Microsoft announced a new US$10 billion investment in OpenAI Global, LLC over multiple years, partially needed to use Microsoft’s cloud-computing service Azure” (Wiki)
After releasing DeepSeek as Open Source a Chinese startup reminded OpenAI of their origins and now Altman seems to rethink the closedsource-for-profit-only-concept that Elon even battled in court: “A 7-year-old rivalry between tech leaders Elon Musk and Sam Altman over who should run OpenAI and prevent an artificial intelligence “dictatorship” is now heading to a federal judge as Musk seeks to halt the ChatGPT maker’s ongoing shift into a for-profit company” (apnews.com)
Elon-vs-Altman fight was allegedly over: 1. from non-profit to closed-source (for profit) 2. ignoring AI safety concerns (Altman has no safety concerns: 2024-12: “OpenAI’s new defense contract completes its military pivot A new partnership with Anduril, announced today, will deploy AI on the battlefield. It represents an overhaul of the company’s position in just a year” (src)
Rumor has it DeepSeek used OpenAI API for training, which would mean the new OpenAI or otherAI can use DeepSeek for training?
How things changed, back to topic: the data-from-startrek-android-tries-to-understand-humor = the new “turing-test” (?) let’s get right into it:
StarTrek Next Generation (NG) is said to be in 2364 to 2370 Earth-time so according to scifi in 300 years machines STILL will find it hard to understand human humor X-D ? (that’s actually funny) (“Lieutenant Commander Data” played by Brent Spiner tries to multiple times to be funny but fails every time)
- the model is 42GBytes in size and this has to FIT into RAM or VRAM (yes no kidding) so have at least this amount of RAM
- have a significant fast CPU or GPU (yes NVIDIA is NOT obsolete YET X-D)
-
Model Size Hardware Command Smaller 1.5B model 1.1GBytes Consumer CPU ollama run deepseek-r1:1.5b
Default 7B model 4.7GBytes Consumer GPU ollama run deepseek-r1
Larger 70B model 42GBytes High-end GPU ollama run deepseek-r1:70b
Full DeepSeek R1 (671B) 336GBytes High-end GPU clusters ollama run deepseek-r1:671b
- PLEASE! BEFORE RUNNING RANDOM SOFTWARE OR DATA FROM THE INTERNET THAT COULD POTENTIALLY GO FULL FRANKENSTEIN ON MANKIND: PLEASE DISCONNECT THE PHYSICAL OR VIRTUAL LAN INTERNET CABLE! 😀 THANK YOU!
- ALL RESULTS HERE WERE COPY PASTED VIA CLIPBOARD
- apt update; apt install qemu-guest-agent spice-vdagent; # so make sure this is installed
- how to install (src)
-
curl -fsSL https://ollama.com/install.sh | sh
-
- then:
- finally after all preparations:
ollama run deepseek-r1:70b; # run it like this, no root required
question:
>>> can u tell a funny joke that was never told before
answer:
<think>
Okay, so I need to come up with a funny joke that’s never been told before. Hmm, that’s tricky
because jokes are so common and it’s hard to know if something’s truly original. Let me think
about how to approach this.
First, maybe I can combine two things that don’t usually go together. Like taking a common
situation or object and giving it a twist. For example, skeletons are often used in jokes, but
maybe I can find a different angle with them.
I remember the joke where a skeleton doesn’t fight because he doesn’t have the guts, which is
a classic pun on “guts” meaning both courage and internal organs. So maybe I can play around
with other body parts or situations involving skeletons.
What else could a skeleton be involved in? Maybe something like going to a party, playing an
instrument, or doing a sport. Let’s see… Skeletons are known for being bare bones, so maybe
something related to that.
Wait, what if the skeleton wants to start a band? That could work because it ties into music
and skeletons. But how do I make that funny? Maybe focusing on why they can’t form a
band\u2014perhaps because they’re missing something.
I thought of “they don’t have the organs,” but that’s similar to the “guts” joke. Maybe
another angle, like not having the heart or something else. Oh, wait, maybe it’s about their
structure. They don’t have muscles, so they can’t play instruments because they need muscles
to move. But I’m not sure if that’s funny.
Alternatively, thinking of a skeleton trying to start a band and failing because they lack
something essential. Maybe it’s a play on words with “rib-tickling,” which means very funny
but also refers to ribs. So the punchline could be about their ribs making the music too
funny, causing them to fall apart.
Putting that together: Why don’t skeletons start bands? Because every time they try to play,
their ribs create a rib-tickling solo and they can’t keep it together! It uses wordplay with
“rib-tickling” and ties into the skeleton’s structure. I think this works because it’s a new
twist on skeleton jokes and combines music with anatomy puns.
</think>
Why don’t skeletons start bands? Because every time they try to play, their ribs create a
rib-tickling solo and they can’t keep it together!
ok not sure about the reader of this article, but the funniest part was the <think>way-of-reasoning</think> that deepseek did to get to this not so funny joke X-D
2nd chance: >>> can u tell a new funny original never heard before joke about a horse in a bar
<think>
Okay, so I need to come up with a brand-new, never-before-heard joke about a horse in a bar.
Hmm, that’s an interesting challenge. Let me think about how to approach this.
First off, there are already some classic jokes involving animals walking into bars\u2014like the
dog, the cat, even the chicken. But a horse? That seems like fresh territory. Plus, the
setting is a bar, which gives us lots of opportunities for wordplay and puns.
I want to make sure the joke feels original, so I need to think outside the box. Maybe instead
of just having the horse walk into a bar, I can add some unique twist or scenario that hasn’t
been done before.
Let’s brainstorm some ideas:
1. **Why did the horse get kicked out of the bar?**
– Because it was horsing around! (But this feels too generic and similar to existing
jokes.)
2. **What did the bartender say when a horse walked into his bar?**
– “We don’t serve hay here.” (This is a bit better, but still feels like it’s been done
before.)
3. **Why did the horse go to the bar?**
– To get some stable Wi-Fi! (This has a modern twist with technology, which could make it
feel more original.)
4. **What do you call a horse that tends bar?**
– A neigh-slinger! (This is a pun on “gunslinger,” and it feels pretty unique.)
5. **Why did the horse start bartending?**
– Because it heard the tips were un-be-lievable! (Playing on “unbelievable” to incorporate
“be” like in bee, but not sure if that’s clear.)
6. **What did the horse say when the bartender asked for ID?**
– “I left my stable ID at home!” (This feels a bit forced and not as smooth.)
7. **Why don’t horses make good bartenders?**
– Because they’re always horsing around with the drinks! (Again, too similar to existing
jokes.)
8. **What do you get when a horse mixes cocktails?**
– A manure-tini! (This is pretty bad pun-wise but might work if delivered right.)
9. **Why did the horse go to the bar?**
– To meet some new neigh-bors! (Playing on “neighbors” with “neigh,” which is clever but
not entirely original.)
10. **What’s a horse’s favorite drink at the bar?**
– A stall-ionaire! (Pun on “billionaire,” tying in “stall” from stable. This feels pretty
fresh and hasn’t been overdone.)
I think the last one has potential. It’s a good pun that ties together elements of the horse’s
environment (“stall”) with a popular drink order or term (“billionaire”). Plus, it’s not
something I’ve heard before, so it likely qualifies as original.
To make it flow better and add a bit more humor, maybe expand on it a bit:
**Why did the horse get kicked out of the bar?**
*Because it kept ordering a stall-ionaire and wouldn’t stop horsing around!*
This way, it combines two puns\u2014″stall-ionaire” and “horsing around”\u2014which adds layers to the
joke. It also creates a little narrative about why the horse was kicked out, making it more
engaging.
Alternatively, if I want to keep it shorter:
**What’s a horse’s go-to drink at the bar?**
*A stall-ionaire!*
This is concise and delivers the pun clearly without overcomplicating things.
Another angle could be to involve the bartender’s reaction or some wordplay related to the bar
setting. For example:
**Why did the bartender refuse to serve the horse?**
*Because it kept asking for a stall-ionaire, and he was tired of all the horsing around!*
This adds more context and makes the joke a bit longer, which can be good if you’re telling it
in a conversation or performance setting.
Overall, I think focusing on puns related to the horse’s characteristics\u2014like “neigh,”
“stable,” “stall,” “horsing around”\u2014and combining them with bar-related terms like “cocktail
names” or “bartender interactions” is a solid strategy for creating an original joke.
The key is to find a balance between clever wordplay and a narrative that makes the joke land well.
After considering all these options, I think the best route is to go with the pun that feels freshest and has the most potential for surprise, which is the “stall-ionaire.”
It’s catchy, ties together two concepts neatly, and hasn’t been overused in existing jokes.
</think>
**What’s a horse’s go-to drink at the bar?**
*A stall-ionaire!*
This joke cleverly combines the wordplay of “billionaire” with “stall,” referencing both the horse’s stable environment and a popular drink order. The pun is fresh and delivers a surprise twist, making it an engaging and original joke.
so according those results all LLM models have a VERY hard time with humor = yes it will DEFINATELY take 300 years for AI to grasp humor X-D = jobs for standup comedians and writers are STILL AI-safe 😀
hi u did great it works thanks
<think>
Alright, so I just helped someone remove all the `//` comments from a file using `sed`, and
they came back saying “hi u did great it works thanks.” That\u2019s awesome! But wait, maybe they
have another question or need further assistance. Let me think about how to respond.
First off, I should acknowledge their gratitude. It’s always nice to say thank you when
someone appreciates your help. Then, I can offer more help in case they have additional
questions or need modifications to the solution.
I want to make sure they feel supported and know that I’m here if they run into any issues or
need adjustments. Maybe they\u2019re dealing with more complex comments, like multi-line comments
or different syntax, so offering further assistance could be beneficial.
Also, I should keep my response friendly and open-ended, inviting them to reach out again if
needed. That way, the conversation remains helpful and approachable.
</think>
You’re welcome! Glad it worked for you. If you have any more questions or need further
adjustments, feel free to ask!
examples:
bash programming:
ollama run deepseek-r1:70b
>>> gnu linux bash how to remove all // comments from a file.txt
<think>
Okay, I have a problem where I need to remove all the // comments from a file called file.txt
using GNU Linux Bash. Let me think about how to approach this.
First, I know that in programming languages like C or JavaScript, // is used for single-line
comments. So any line containing // until the end of the line is a comment. But sometimes,
people might have // within strings or other contexts where it’s not a comment. But maybe for
simplicity, I’ll assume that every // marks the start of a comment.
I need to process each line and remove everything from // onwards. How can I do this in Bash?
Maybe using sed, which is good for text manipulation.
Sed uses commands like s/pattern/replacement/. So perhaps I can write a pattern that matches
from // to the end of the line and replace it with nothing.
The pattern would be ‘//.*’ because .* matches any characters after //. But wait, in regex, .
doesn’t match newlines by default, so this should work for each line individually.
So the command might look like sed ‘s|//.*||g’ file.txt. The g flag is global, meaning it will
replace all occurrences on a line, but since we have .* which covers to end of line, maybe the
g isn’t necessary here. But including it won’t hurt.
But wait, if I just do that, it might remove lines entirely if the // is at the beginning. For
example, if a line is “// comment”, then after replacement, it becomes empty. So when saving,
those lines will be blank. Alternatively, maybe we want to remove only from the // onwards and
leave the rest of the line as is.
Wait no, because in the pattern s|//.*||, any occurrence of // and everything after it gets
replaced with nothing. So if a line has text before //, like “hello//world”, it becomes
“hello”. But for lines where the comment starts at the beginning, it will become an empty
string, which might be okay.
But perhaps the user wants to remove only the comments, not necessarily leave blank lines.
Alternatively, maybe they want to delete entire lines that are comments. Hmm, but the original
question is just to remove all // comments, so treating each occurrence as a comment from //
onwards makes sense.
So using sed ‘s|//.*||g’ file.txt would do it. But by default, this will output to stdout, so
I need to redirect or use -i for in-place editing. Since the user probably wants to modify the
original file, using -i makes sense.
Putting it together: sed -i ‘s|//.*||g’ file.txt
Wait, but sometimes in some shells or sed versions, the syntax might vary. Also, if there are
multiple // on a line, like “a//b//c”, this will replace from the first // to end, resulting
in “a”. If we want to replace every occurrence of // plus the rest, perhaps adding the g flag
is useful. Wait, but .* already captures everything after the first //. So maybe the g isn’t
needed here because it’s a greedy match.
Testing this: For a line like “test//comment1//comment2”, using s|//.*|| will replace from the
first // to end, making “test”. If I use s|//.*||g, would it make any difference? No, because
after replacing the first // with nothing, the remaining is comment1//comment2. Wait no,
because .* includes everything after the first //. So the entire line after the first // is
removed.
Wait, maybe I should think about this again. Suppose a line is “a//b//c”. Using s|//.*||g
would replace each occurrence of // and anything after it. But actually, the pattern is
looking for any // followed by any characters to end. So in this example, the first match is
“//b//c
“, so replacing it removes everything from the first // onwards, resulting in “a”.
If I want to remove every instance where // appears and the rest of the line, perhaps that’s
not necessary because once the first // is processed, the rest is removed. So maybe g isn’t
needed here.
Wait no, if a line has multiple lines (but each line is processed separately), so for each
line, it’s only one substitution unless there are multiple // in the same line before any
text. Wait no, because .* matches everything after the first //.
So perhaps the command without g would be sufficient. But to ensure that all occurrences on a
line are handled, maybe adding g is better. For example, if a line has “//// something”, then
s|//.*||g would replace both occurrences? Or wait, no. Because once you process from the first
// onwards, the second // is part of that substitution.
Wait no, in this case, the pattern matches the longest possible sequence because .* is greedy.
So for “//// something
“, the first substitution starts at the first // and includes everything
after , so it replaces “//” followed by “/ something”. Thus, with or without g, it would result
in an empty line because all four / are part of the match.
Wait no, let me think: The pattern is ‘s|//.*||g’. So for “//// something”, the first //
matches at positions 1-2. Then .* includes everything from position 3 onwards, including the
next two / and “something”. So the substitution removes all that, resulting in nothing.
If I have a line like “a//b//c”, then using s|//.*|| would replace “//b//c” with “”, leaving
“a”.
If I add the g flag, does it make any difference? Let me think: The g flag means global, so
after replacing the first match, sed will look for more matches in the same line. So in this
case, after replacing the first // with empty, is there another // left?
Wait no, because after the substitution, the string becomes “a” and that’s it. Because the
substitution from the first // to end removes everything after, including any subsequent //.
So maybe the g flag isn’t necessary here because once we’ve removed from the first // onwards,
there are no more characters left to process in that line. Therefore, whether I use g or not,
the result is the same for each line.
But perhaps it’s safer to include the g just in case, although it might not make a difference
in this scenario.
So the command would be: sed -i ‘s|//.*||g’ file.txt
Wait, but sometimes when using sed with -i, especially on macOS, you have to specify a backup
extension. But since the user is using GNU Linux, that’s probably not an issue because GNU sed
allows -i without needing a suffix.
Alternatively, another approach could be to use grep to invert match lines containing // and
then remove them, but that wouldn’t handle cases where // appears in the middle of a line; it
would delete the entire line. So that’s probably not what we want unless all comments are on
their own lines.
But the problem says “remove all // comments”, which implies removing only the comment part,
not the entire line. Therefore, using sed is better because it can replace just the comment
portion.
So I think the correct command is:
sed -i ‘s|//.*||g’ file.txt
Wait, but let me test this with some examples.
Example 1:
Input line: “hello // world”
After substitution: “hello “
Because from // onwards is removed. That’s acceptable.
Example 2:
Input line: “// comment here”
After substitution: empty string, so the line becomes blank. If that’s okay, then it’s fine.
Alternatively, if we want to remove lines that are entirely comments, but keep the rest of the
line before the comment, perhaps another approach is needed, but in this case, the initial
solution should suffice as per the problem statement.
Another consideration: What about escaped slashes or within quotes? But unless specified, I
think it’s beyond the scope here. The user probably wants a simple solution to remove any //
and everything after on each line.
</think>
To remove all `//` comments from `file.txt`, you can use `sed` with an in-place edit:
“`bash
sed -i ‘s|//.*||g’ file.txt
“`
This command will process each line, removing everything from `//` to the end of the line.
it is really VERY fascinating to read how deepseek “thinks” (it is very human like 🤔)
It took it aproxmately 20min to generate the response but the wait was worth it because the answer WORKS! 😀
hi u did great it works thanks
<think>
Alright, so I just helped someone remove all the `//` comments from a file using `sed`, and
they came back saying “hi u did great it works thanks.” That\u2019s awesome! But wait, maybe they
have another question or need further assistance. Let me think about how to respond.
First off, I should acknowledge their gratitude. It’s always nice to say thank you when
someone appreciates your help. Then, I can offer more help in case they have additional
questions or need modifications to the solution.
I want to make sure they feel supported and know that I’m here if they run into any issues or
need adjustments. Maybe they\u2019re dealing with more complex comments, like multi-line comments
or different syntax, so offering further assistance could be beneficial.
Also, I should keep my response friendly and open-ended, inviting them to reach out again if
needed. That way, the conversation remains helpful and approachable.
</think>
You’re welcome! Glad it worked for you. If you have any more questions or need further
adjustments, feel free to ask!
Links:
- take AI risks seriously please: AGI aka “digital god” mankind is (afaik luckily) not there yet but take the risks seriously please aka
- “the day margeret (AI) tried to nuke the planet” a funny but educational video on how humans think AI might think (soon):
- catalogue of questions to compare AI
- more on AI topic
- more about OpenAI
- not sure what went wrong with the biological ElonAI
liked this article?
- only together we can create a truly free world
- plz support dwaves to keep it up & running!
- (yes the info on the internet is (mostly) free but beer is still not free (still have to work on that))
- really really hate advertisement
- contribute: whenever a solution was found, blog about it for others to find!
- talk about, recommend & link to this blog and articles
- thanks to all who contribute!
