
Deepseek R1 is out - available via Deepseek API or free Deepseek chat. If you are following LLM/Gen AI space you've likely seen titles, read posts,...
Some comments have been hidden by the post's author - find out more
For further actions, you may consider blocking this person and/or reporting abuse
I performed a moderately complex, real-life query against the usa stock market using ChatGPT and DeepSeek. I only had one measurement of success: an accurate return of information. ChatGPT did extremely well - it added a chart (not requested) and much more information than just the closing price; in other words, about what you get on Yahoo Finance stock queries. DeepSeek merely replied with "that information isnt in my system yet". I realize that DeepSeek is a tool in its relative infancy, and in time it will be just as powerful and useful as ChatGPT. But as of now, it is useless to me.
Query: what is the stock market price of AAPL
I used that same query in deepseek and this is the answer I get:
As of the latest available data in the search results (dated January 28, 2025), the closing stock price of Apple Inc. (AAPL) was $238.26, reflecting a +3.65% increase from the previous trading session . During extended trading hours on January 28, the price slightly fluctuated to $238.34 (+0.04%) . Pre-market data for January 29, 2025, shows a slight decline to $237.34 (-0.39%) as of 4:29 AM EST .
Key Details:
Recent Performance:
Market Context:
Upcoming Catalyst:
For real-time updates, refer to financial platforms like Yahoo Finance or Google Finance.
sounds stupid but... are you sure you activate web search?
What I've experienced is that if you ask something Political regarding China especially, you'll observe that it gives an a Biased Answer.
Whereas for CHAT GPT, even if you ask the same for America, observe neutral replies.
I think it won't last long if it stays Biased towards a country or developer.
As a user I am annoyed. The Deepseek server is always busy.
^^^^ THIS
Better performance /= stability
I have the same problem with "always busy"
Host it locally :-) No more issues
Thanks a lot :). I am also disappointed by the quality of the answers as compared even to GPT-o4
Glad to know that I'm not the only one lol
My first thought was: why would anyone even try to play chess with LLM? There are better "algorithms" for that.
But yes, especially in todays world where everybody is expecting AGI with every new LLM why it shouldn't play chess after all?
Where I see the problem (and a limitation) of current approach is the misuse of general purpose reasoning capabilities of LLMs, which are undeniably there but only in emergent form. What would average person do if confronted with this task? Most of us would just use some software, smaller portion would implement their own software and only handful (of chess masters) would play by themselves.
Unless we equip these 'AI' models with complete set of tools, environment and ability to use workflow patterns to model and execute a workflow designed for a given task, results are going to be quite disappointing, at least for the nearby future.
For me currently, DeepSeek has been extremely unhelpful and annoying. The servers are always busy. I understand that DeepSeek is still in its early stages, but the developers should have tested their product before publishing it. In my personal opinion, ChatGPT is actually the best tool to use professionally, and it has become my favorite. I do hope that DeepSeek continues to advance though. Maybe some day, DeepSeek will be just as powerful, if not more powerful, than ChatGPT. Cheers to a promising future!
If the only complain is Deepseek's web app availability (not the quality of responses) - it's hard to blame the folks for not burning millions to serve the avalanche of user requests, for free. Don't you think so?)
This isn't the only issue, though. One of the biggest issues that I myself and every other user have is that the responses in DeepSeek are way less intricate than ChatGPT. I am sorry, but DeepSeek just isn't cutting it. The biggest issue is that DeepSeek isn't as secure as the developers say it is because multiple users have already had hackers obtain their credit card info when the users have paid for the subscription. I myself did a penetration test (specifically a SQL Injection and changing admin cookie values from False to True in DeepSeek) and sure enough, all credit card info from the credit cards that users have used to pay for the subscription was listed. As someone who is a Network Cybersecurity Architect, this extremely worries me because this was the first attack that I thought of. An SQL Injection is always the first attack that hackers will think of and try. This is the exact reason why I haven't paid for a subscription myself. There are so many more issues with DeepSeek that I can't list them all. The DeepSeek developers shouldn't have released the product without doing penetration testing on it themselves. Sorry if you think differently, but at this point the poor performance and the vulnerabilities of DeepSeek have become too much of a problem to ignore. Professionally speaking, if the developers want to save their product, the only chance that they have is if they unpublish their product, do a full on remediation, do their own penetration testing of all possible parameters, and republish the product.
Can you share more detail about SQL injections that you tried with DeepSeek (presumably chat.deepseek.com)? What, when, how..
It seems DeepSeek has many great talents in AI development/optimization, but doesn't care much about good web developers. The way how it pops up lots of Cloudflare CAPTCHA boxes all across the website at times shows they are not even trying to make good quality web frontend and backend...
I tried the SQL Injections on January 31, 2025. I just used SQLite to automatically connect to one of DeepSeek's servers, and it actually worked. And yes, it was chat.deepseek.com
i think the DeepSeek folks are just lacking good web developers, unlike all the brilliant AI developers they have. They really need to recruit some better web developers, the way how dozens of Cloudflare CAPTCHA boxes can pop up in a single page just shows bad web developers at work
That is the least of their worries. Really, they need to actually hire some trustworthy folks that can actually provide good security for all of the sites. In my opinion, they should focus on security first, then they can start fixing the product itself.
Nice post.. Had a question for you after seeing this. How good is deepseek than others w.r.t hallucinations, as yesterday night I was reading a content piece on LLM hallucination detection..
Assuming that a hallucinations include cases when the model does not what it is asked for (e.g. skips an acton) or decides to make move that is not legal - R1 is quite bad. You can tell this by large number of mistakes and low average game duration (you can see that by hovering over a row in the leaderboard)
To be honest, I'm just happy there's a contender for "best AI model" that isn't made by a multi billion dollar tech corporation from the US (well, Deepseek's company might be worth billions now lol). All that being said, for the few times that I actually utilize AI in my own work, I still use ChatGPT.
This is such an interesting topic! I'm really happy to hear about it. I absolutely Love it, and you should definitely check out more topics like this.