cs hashtag performance
The TikTok hashtag #cs encompasses a variety of content related to computer science, coding, tech tips, programming challenges, educational tutorials, student experiences, career advice, innovative projects, and geek culture.
551.0k
What if I told you that building a world-class AI like ChatGPT no longer costs $100 million? Meet DeepSeek V3, the AI model thatās rewriting the rules of whatās possible. With just $5.5 million in training costs and some incredible engineering, it outperforms big names like GPT-4 and Claude 3.5 on tasks like coding, math, and long-text comprehension. Hereās the wild part: DeepSeek isnāt just cheaperāitās smarter. Instead of using its entire brain (617 billion parameters), it activates only 37 billion parameters for any task, making it 10x more efficient while staying just as powerful. It also predicts two words ahead instead of one, speeding up its responses and improving accuracy. How did they do it? Through clever innovations like: ā¢ DualPipe: A training trick that lets their systems learn and work simultaneously. ā¢ High-Quality Data: Instead of scraping the entire internet, they curated precise, multilingual training data thatās focused on what mattersāEnglish, Chinese, math, and coding. And hereās the kicker: Itās open source. Anyone can access the model, tweak it, or even build their own version. This changes the game. Startups, universities, and even solo developers can now compete with AI giants like OpenAI and Meta. This isnāt just about a single model. DeepSeek V3 is showing the world how accessible and affordable cutting-edge AI can be. The future of AI isnāt just biggerāitās smarter and more open than ever. Hashtags #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #deepseek #chatgpt #v3 #opensource #revolutionary
352.0k
AIās Memory Problem Today, Iām diving into one of the most pressing challenges for large language models (LLMs): memory. These models are incredibly intelligent, but their memory is shockingly short. Think of it as having a genius assistant who forgets your last conversation after 10 minutes. The best models today, like O1 Pro, boast 200,000 tokens of memory, but even thatās just a fraction of whatās needed for meaningful, long-term interactions. Hereās the kicker: solving this isnāt cheap. I crunched the numbers in my Substack, and providing even basic long-term memory for a platform like ChatGPTāserving 125 million daily usersāwould cost over half a trillion dollars annually. And thatās not for human-level memory; thatās just a few months of retention. The real question is: how do these memory limits shape what kinds of problems we solve? What are we inherently limited from tackling because our AI thinking partners canāt remember beyond a few exchanges? Even more fascinating, how does this shape our memory? Thereās anecdotal evidence of people adjusting their thinking and vocabulary because of how they interact with LLMs. Are we adapting to their limits, or will these limitations drive the next breakthrough in AI? I believe memory will dominate AI discussions in 2025, and itās a topic we need to address now. If this resonates, check out my Substack for a deeper dive, and letās keep the conversation going. Hashtags: #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #2025 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #chatgpt #memory #question #future
342.6k
AWPer of the Year š 2nd Ranked Player š„ THATāS OUR GOAT! #hltv #cs #esports #m0nesy #counterstrike
288.2k
Whatās your take on the news here? Thereās SO MUCH Hereās the breakdown of whatās happening in the world of artificial intelligence. First, thereās a case connected to OpenAI thatās sparked important discussions about the future of the fieldādefinitely one to watch. Next, NovaSkyāa startup no one saw comingādropped its SkyT1 model, a 32-billion parameter system that matches OpenAIās o1Preview. The kicker? It costs only $450, making it clear that AI capabilities are becoming way more affordable. OpenAI is doubling down too, hiring robotics engineers to dive into robotics. I predicted this move, and itās exciting to see it confirmed. Even more surprising? OpenAI is paying $385K salaries for React engineers, showing that human expertise still plays a huge role, even with AI coding capabilities. Meanwhile, Hyperbolic Labs announced that AI agents are already renting GPUs to code in PyTorch. This shiftāmachines running their own tasksāhappened faster than anyone expected. And if youāre frustrated with how ājudgyā some AI like Claude can be, users are now sharing hacks to make interactions smoother and more productive. On the creative side, Sunoās founder sparked a heated debate by claiming that making music is too hard and not enjoyable. Musicians clapped back, saying that effort is the soul of creation. Grok, an AI from X, saved the day in a medical setting, diagnosing a broken wrist that urgent care missed. Thanks to Grok, the family avoided surgery with a quick intervention. Finally, energy concerns? Running 80,000 tokens on a large AI model uses about as much energy as an hour of Netflix streaming. But with AI models becoming more efficient, the energy conversation is evolving too. AI is moving faster than everāwhatās your take on this weekendās news? Hashtags: #ai #aiupdates #openai #novasky #product #productmanager #productmanagement #startup #business #llm #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #2025 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt
279.5k
āHow to smallest hitbox in cs2ā #cs2 #csgo2 #counterstrike2 #cs2funny #csgo #Cs #counterstrike #cs2fails #cs2update
270.6k
Wildest AI Internship Ever This might be the craziest AI story Iāve ever heard. Meet Keyu Tian, an intern at ByteDance (the company behind TikTok), who hacked his way into GPU access to produce groundbreaking AI research. Back in mid-2024, Tian joined ByteDance, but things went downhill fast. Training pipelines started failing, models mysteriously crashed, and tiny edits sabotaged workflows. All this chaos freed up GPUs, which Tian secretly used for his own research. ByteDance eventually caught on, fired him in August 2024, and even reported him to his university. But Tian didnāt stop there. In October, he submitted his paper, Visual Autoregressive Modeling: Scalable Image Generation via Next Scale Prediction, to NeurIPSāthe worldās most prestigious AI conference. And guess what? It won Best Paper in December 2024. Hereās why the paper is a big deal: it pushes the boundaries of image generation, going beyond token or pixel predictions to predict larger image scales, making the process faster and more efficient. The award sparked controversy because Tian used stolen GPU time to do the work. Meanwhile, ByteDance is suing him for $1.1 million in damages. Where do I stand? Tian is undeniably brilliant but also reckless. If he doesnāt get hired and managed properly, someone will still benefit from his ingenuity. This story is far from over, but itās already a wild rideāpart genius, part chaos, and 100% unforgettable. #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #bytedance #intern #crazy #real #story #wild #true #strangertruth #weird #future #2025
122.2k
AI Career Strategy 2025 2025 is here, and itās time to rethink how we approach career growthābecause AI is changing everything. If youāre still relying on the traditional job family mindset, itās time to adapt. The next decade isnāt about sticking to old paths; itās about agency and learning how to leverage AI to expand your potential. Hereās my strategy: I use an AI Integration Framework to map my expertise in three layers: 1. Core Expertise: The areas Iāve mastered 100%. 2. Secondary Knowledge: Areas where AI can elevate me to the top 1%. 3. Extensions: Skills Iām 40-60% confident in, but where AI can take me to the next level. This approach transforms a traditional T-shaped skillset into a multi-T-shaped career, with depth in multiple areas and the breadth to navigate new challenges. The risk? Staying stuck in outdated mindsets. The opportunity? Building an adaptable career where AI amplifies your unique talents. Hereās what Iām doing now: ā¢ Mapping my skills into the framework. ā¢ Expanding aggressively into areas where AI adds the most value. ā¢ Owning my career trajectory by staying ahead of the curve. Whatās your AI learning plan for 2025? Comment below and letās discuss and build together. Donāt let this year pass without leveraging AI to define your future. #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #careercoach #careerguidance #2025 #goal #resolution #newyear #agi
114.2k
Who should be more worried, Google or ChatGPT? The recent tragedy in Las Vegas brought up a surprising angle: how people are using AI tools like ChatGPT instead of traditional search engines like Google. Letās dive into the tech side of this story because the real shift here isnāt about AIās āmagic powersā but how search behavior is evolving. When you look at the queries that were reportedly made, they werenāt complex or unique to large language models (LLMs). They were short, straightforward questionsāthings Google has been handling for decades. But whatās telling is this: the person involved chose ChatGPT over Google. Why? It was faster, more direct, ad-free, and easier to use. This isnāt a story about AI doing something new or extraordinary. Itās a story about how conversational AI is quietly replacing traditional search engines for simple, ābits and piecesā queries. It shows how much we value instant, clean answers over digging through pages of search results. From a tech perspective, this also highlights gaps in AI safety. Current guardrails focus on blocking malicious or complex queries, but they donāt catch these incremental, seemingly benign ones. This shift in behaviorāfrom Googling to āChatGPTingāāis something everyone in tech and AI needs to think about. The real takeaway? This isnāt just about ChatGPT or Google. Itās about how technology shapes behavior and where weāre headed with tools that are becoming second nature in our daily lives. #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #chatgpt #google #real #story #future #2025 #cursor #replit #pythagora #bolt
107.1k
Itās not just you! And yes it happened again š OpenAI's services crashed yet again today, with ChatGPT and their video generation tool Sora experiencing a major outage on December 26. This marks the second significant disruption in just two weeks, following their previous system-wide failure earlier this month. As someone deeply embedded in AI development, watching these repeated outages is concerning. The disruption began early morning after Christmas, leaving millions of users unable to access these critical AI tools right as they returned to work. What's particularly striking about this outage is the concurrent failure of both ChatGPT and Sora, mirroring the pattern we saw in the previous crash. The timing is especially problematic - hitting during the post-holiday return to work when businesses rely heavily on these AI tools to manage their backlog. Having worked with these systems extensively, I can say these recurring outages point to deeper infrastructure challenges at OpenAI. This isn't just about temporary downtime; it's about the reliability of AI services that have become essential to modern workflows. The repeat nature of these failures raises serious questions about system stability and the need for more robust failover protocols. #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #outage #systemdown #aidown #repeatoutage #systemfailure #techissues #downagain #reliabilityāāāāāāāāāāāāāāāā
106.2k
Remember the internet in 1995? Most people didnāt get it yet. It was clunky, slow, and seemed like a niche thing for nerds. Fast forward a few years, and suddenly we couldnāt live without it. Thatās where we are right now with AIābut instead of taking years to evolve, itās happening in months. AI isnāt some sci-fi fantasy or overhyped trend. Itās software that can thinkānot like a human, but in a way thatās crazy fast and super powerful. A year ago, it could help write emails or give decent answers to questions. Now? Itās solving advanced math problems, writing professional code, and automating tasks that used to take whole teams to do. Whatās wild is how quickly this is evolving. Imagine if, in 1995, the internet went from slow dial-up to Netflix streaming in two months. Thatās AI right now. By 2025, these systems will be able to handle entire workflowsābooking trips, managing schedules, or even running parts of a businessāall based on simple instructions. And hereās the kicker: you donāt need to be a coder to use it. AI is getting so good at understanding plain language that anyone will be able to build apps or solve problems just by describing what they want. Itās not just a tech thingāitās going to touch every job, every industry, and every part of how we live. So yeah, itās moving fast, and itās a little overwhelming. But just like the internet back in the day, itās not about whether AI will be a part of the futureāitās already here. The question is: how will you use it? Hashtags: #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt #chatgpt #memory #question #2025 #explain #history #agi
92.0k
MCSGA - Make CS Great Again (procrastinated my annual check up till the last day of the year š) #CS #compsci #computerscience #frankniu #tech #jobmarket #student #vexbolts
84.4k
āļøš #sound #music #cs2 #cs #2025 #newyear #happynewyear #christmas #jingle #jingles #jinglebells #bells
69.2k
AI Predictions for 2025 When you and I write New Yearās reflections, nobody pays attention. When Sam Altman does, the entire world leans ināand for good reason. Heās a founder of OpenAI and co-creator of ChatGPT. Samās 2025 vision includes workplace AI agents collaborating in tools like Slack. While he didnāt discuss pricing, Iād guess they could cost around 10% of a human employeeās salary based on similar models. These AI tools are designed to thrive in partnership with humans, not replace them, emphasizing collaboration over automation. Altman expects broadly distributed benefits from these tools, which suggests jobs are safe for nowāat least in environments where strong human-AI teamwork exists. On a broader scale, OpenAIās strategy is split between refining products like ChatGPT and diving deep into research on artificial superintelligence (ASI). Altman sees ASI as a tool to amplify humanityās ability to tackle complex problems, not a force to replace us. OpenAI is prioritizing scientific applications like curing aging, treating diseases, and advancing cancer research. Altmanās vision for ASI is rooted in solving fundamental challenges, making it clear that he views AI as a partner in progress, not a disruptor of human agency. Altman doesnāt predict immediate workplace disruption but believes AI agents will transform workflows as trusted collaborators. For me, this aligns with a prediction Iāve made before: the future is less about AI taking over and more about smarter tools enhancing how we work. The shift might feel revolutionary, but at its heart, itās all about creating better human-AI teams. What do you think? Ready to team up with an AI coworker? Let me know in the comments! #product #productmanager #productmanagement #startup #business #openai #llm #ai #microsoft #google #gemini #anthropic #claude #llama #meta #nvidia #career #careeradvice #mentor #mentorship #mentortiktok #mentortok #careertok #job #jobadvice #future #2024 #2025 #story #news #dev #coding #code #engineering #engineer #coder #sales #cs #marketing #agent #work #workflow #smart #thinking #strategy #cool #real #jobtips #hack #hacks #tip #tips #tech #techtok #techtiktok #openaidevday #aiupdates #techtrends #voiceAI #developerlife #cursor #replit #pythagora #bolt
63.7k
Which Back Taper You Rocking With ?š #fypć· #viral #foryou #foryoupage #trending #relate #parati #barbertok #mexicantiktok #taper #2freshcs #cs
start an influencer campaign that drives genuine engagement
Technology
Fastest Growing