NEET
09/25/2026 (Fri) 20:55
No.1034047
del
09/25/2026 (Fri) 20:55
No.1034047
del
>>1034037
Also different models, how do they perform for cost and detection measures will need to be assessed. Executive around are going to ask, why are we not using Elongs AI compared to OpenAI. A junior analyst could do that work supervising it and evaluating it. If you are doubling cost for the new meme model, are you seeing the benefit from it should be assessed. There are ways to do it cheaply like using cheaper models but again you would need someone to supervise that process for reporting and manage it all
Also given different log sources what do they mean and can the AI understand what it all means when put together, do we care about this feature and does it solve a problem efficiently. Some AI of them are good in like Microsoft suite for their products but if I dump everything into a single Syslog table, does the AI understand how everything is transformed and how it all connects together, I have my doubts. From what I have seen and learned AI is great at doing what it knows and what it has been trained on, but outside those tasks it can be a lot weaker. Until that uplifts for AI agents, I will remain skeptical and isnt cheap, every vendor offers it but be prepared to pay.
The use of AI in cybersecurity is still very much beneficial to create queries and analyse, but again it is all directed by a human who has expertise on what looks good and bad based on their understanding of the environment.
Also different models, how do they perform for cost and detection measures will need to be assessed. Executive around are going to ask, why are we not using Elongs AI compared to OpenAI. A junior analyst could do that work supervising it and evaluating it. If you are doubling cost for the new meme model, are you seeing the benefit from it should be assessed. There are ways to do it cheaply like using cheaper models but again you would need someone to supervise that process for reporting and manage it all
Also given different log sources what do they mean and can the AI understand what it all means when put together, do we care about this feature and does it solve a problem efficiently. Some AI of them are good in like Microsoft suite for their products but if I dump everything into a single Syslog table, does the AI understand how everything is transformed and how it all connects together, I have my doubts. From what I have seen and learned AI is great at doing what it knows and what it has been trained on, but outside those tasks it can be a lot weaker. Until that uplifts for AI agents, I will remain skeptical and isnt cheap, every vendor offers it but be prepared to pay.
The use of AI in cybersecurity is still very much beneficial to create queries and analyse, but again it is all directed by a human who has expertise on what looks good and bad based on their understanding of the environment.