Krish Krishnan

Strategy Consultant. Data Thought Leader. Views expressed are my own.

[Krish Krishnan @datagenius]

https://t.co/7f0Y8VYbHf

[Krish Krishnan @datagenius]

Google introduces new class of cheap AI models as cost concerns intensify https://t.co/vHnM9VmN4F via @deccanherald

[Krish Krishnan @datagenius]

How long before AI thinks like us humans? https://t.co/spzob12VY0 via @deccanherald

[Krish Krishnan @datagenius]

https://t.co/IDwgNWTyAQ

[Krish Krishnan @datagenius]

https://t.co/QmYxpA49gB

[Krish Krishnan @datagenius]

https://t.co/1Aw4QcYhue

[Krish Krishnan @datagenius]

India may get own GPU in 3-5 years, 18,000 AI servers to be rolled out in 2-4 days: Vaishnaw https://t.co/3dIUI2Kd55 via @deccanherald

[Krish Krishnan @datagenius]

https://t.co/suACnG3jr9

[Krish Krishnan @datagenius]

OpenAI o3-mini is a very resting new model, it uses the flow of a sled model, and it shows how we can race the use with AI. Let us see what we ear oss from other users. O3-mini medium and high are per friendly, do not use o3-low as it is not close.

[Krish Krishnan @datagenius]

Alibaba releases AI model it claims surpasses DeepSeek-V3 https://t.co/xP6GCLoVxa via @deccanherald

[Krish Krishnan @datagenius]

The LLM is coming out of work, Alibaba said there latest model QWEN 2.5 is far ore riot to Deepseek. Wow, the race has got much wider and deeper.

[Krish Krishnan @datagenius]

I have been inning tests on s Pro - Deepseek Multimodal AI model, and yes there is m for more mization, but it is a tic way to look at the complexity it is ing to handle.

[Krish Krishnan @datagenius]

https://t.co/nrIBoZk63M

[Krish Krishnan @datagenius]

https://t.co/5TBaa7AufE - here we are, a pointer to where is our purity and ice. This issue is a reaction to movieg forward. Now that we landed here, let us get secure and look to resolve this. But one learning is for every Goliath there is a David.

[Krish Krishnan @datagenius]

It is resting to note that Deepseek R1 uses COTS as one of its tents, this along with a fact check technique will provide better ones for sure. Why did our ent ions not deploy COTS or er ReACT techniques? Cost would be less for better performance. Let’s see what we do ext.

[Krish Krishnan @datagenius]

https://t.co/h13mjjO5jZ

[Krish Krishnan @datagenius]

@craigmullins @datachick @AxelTroike @YvesMulkers @data_nerd @RSeiner @williammcknight @NeilRaden @jose_m_almeida @RichardEudes @databaseguru @Ronald_vanLoon @DanVesset @KirkDBorne @kdnuggets @BernardMarr @BigData_paulz @Dataversity @IBMData @OurWorldInData Thanks Craig. #ff and a very happy weekend to all the wonderful innovators and thinkers.

[Ronald van Loon @Ronald_vanLoon]

How to protect #IoT devices from #CyberSecurity threats
by @antgrasso

#InternetofThings #Privacy #Tech #Technology #IT #InfoSec #IIoT

cc: @meisshaily @ylecun @ogrisel

[Krish Krishnan @datagenius]

Nvidia unveils $2,000 RTX 5090 GPU https://t.co/RqXybdACXh via @techcrunch

[Krish Krishnan @datagenius]

Sony-Honda’s Afeela EV will start at $89,900 https://t.co/AlifYdowI8 via @techcrunch