- Everyday AI
- Posts
- How To Build an AI-First Company
How To Build an AI-First Company
DeepSeek shocks AI stock market, DeepSeek Dethrones ChatGPT in App Store and DeepSeek releases new AI image model, Alibaba's new AI can control PCs and more!
Everyday AI
January 27, 2025
š Subscribe Here | š£ Hire Us To Speak | š¤ Partner with Us | š¤ Grow with GenAI
Outsmart The Future
Today in Everyday AI
7 minute read
š Daily Podcast Episode: AI can make it hard to compete as a company but itās also the thing that can put you on top of the competition. So how can you implement AI to create an AI-first company? We break it down. Give it a listen.
šµļøāāļø Fresh Finds: Marc Andreessenās thoughts on DeepSeek, Apple iOS AI updates and DeepSeek restricts app signups. Read on for Fresh Finds.
š Byte Sized Daily AI News: DeepSeek shocks AI stock market, DeepSeek dethrones ChatGPT in App Store and DeepSeek releases new AI image model. For that and more, read on for Byte Sized News.
š AI In 5: This new Gemini feature lets you talk to your PDFs inside Google Drive. Is it any good? See it here
š§ Learn & Leveraging AI: Wondering how to build an AI-first company successfully? We break down what you need to know. Keep reading for that!
ā©ļø Donāt miss out: Did you miss our last newsletter? We talked about Stargate Project funding woes, AI achieves self-replication, Metaās $60 billion AI boost, Altman's World Project AI agent plans. Check it here!
How To Build an AI-First Company š¤
One side of the coin ā AI can make it VERY hard to compete when starting a new company.
Other side of the coin ā AI can make it MUCH easier to compete when starting a new company.
Rajesh Kandaswamy knows the path.
After spending a decade as a VP and researcher at Gartner, Rajesh learned a few things about building an AI-native company.
He shares the ins and outs of what it takes to build an AI-first company from the ground up.
Join the conversation and ask Jordan questions on AI here.
Also on the pod today:
ā¢ AI First Approach š¤
ā¢ Overcoming AI Challenges š
ā¢ Future of Organizations and AI Agents šµ
Itāll be worth your 30 minutes:
Listen on our site:
Click to listen
Subscribe and listen on your favorite podcast platform
Listen on:
Hereās our favorite AI finds from across the web:
New AI Tool Spotlight ā Polymet is an AI product designer, Magic Inspector is an AI web test automation platform and Nuvio provides AI-powered financial management.
Trending in AI ā Marc Andreessen warns that DeepSeek is āAIās Sputnik moment.ā
Apple ā Apple has released iOS 18.3 with updates to its AI notification summaries.
Big Tech - Apple and Google were originally worried that Character.ai was inappropriate for teens before Googleās $2.7B deal with the company.
AI Models - DeepSeek is restricting sign-ups to its app after malicious attacks.
Read This ā Brazil has banned Sam Altmanās World from capturing iris scans from its citizens.
1. DeepSeek's AI Surprise Stuns Stock Market š
In a dramatic turn of events, tech stocks plummeted on Monday as Chinese startup DeepSeek launched a cost-effective AI assistant, threatening established giants like NVIDIA. According to Reuters, DeepSeek's rapid rise in popularity led to a significant sell-off, with NVIDIA's shares dropping over 17%, marking a historic one-day loss.
This upheaval in the AI sector could revolutionize accessibility, potentially benefiting small businesses and tech innovators. Despite the market chaos, investors view this as a chance to invest in high-quality tech stocks at reduced prices.
2. DeepSeek Dethrones ChatGPT in App Stores š
DeepSeek, a Chinese AI lab's mobile app, has surged to the No. 1 spot in global app stores, surpassing ChatGPT. This rapid climb to the top is attributed to DeepSeek's release of cost-efficient AI models that challenge industry giants like OpenAI and Google, causing a stir in the tech world and even impacting NVIDIAās stock.
With 2.6 million downloads across platforms, DeepSeek has demonstrated that the consumer AI space is ripe for disruption. However, its climb may face hurdles, as the company reported malicious attacks that have temporarily restricted user registrations outside China.
3. DeepSeek's Janus-Pro: AI's New DALL-E Contender š„
According to DeepSeek, the newly released Janus-Pro models are setting the AI world abuzz by outperforming OpenAI's DALL-E 3 on benchmarks like GenEval and DPG-Bench. These models, ranging from 1 billion to 7 billion parameters, are touted for their compact size and high performance, allowing them to analyze and create images with impressive flexibility.
Available under an MIT license on Hugging Face, Janus-Pro's arrival not only signals a potential shift in AI innovation but also raises questions about the U.S.'s lead in the AI race, as noted by Wall Street analysts.
4. Alibaba's Qwen2.5-VL Can Control PCs and Phones š§āš»ļø
Alibabaās Qwen team has unveiled Qwen2.5-VL, a new set of AI models that outshine competitors like OpenAI's GPT-4o and Google's Gemini 2.0 Flash in tasks ranging from video analysis to document parsing, according to TechCrunch's Kyle Wiggers. Available for testing on Alibabaās Qwen Chat app and Hugging Face, the models can analyze complex data and recognize popular media IPs, although they have restrictions on politically sensitive topics.
Notably, Qwen2.5-VL can interact with both PC and mobile software, though its real-world application might be limited, as shown by its performance on the OSWorld benchmark.
5. Meta's AI Chatbot Gets a Memory Boost š§
Meta is rolling out a new feature that allows its AI chatbot to remember details like your dietary preferences and interests, using data from Facebook and Instagram to tailor recommendations. According to a company blog post, this memory function will become available across Facebook, Messenger, and WhatsApp in the US and Canada, aiming to enhance personalization.
The AI will adjust interactions based on contextāso if you mention being vegan, itāll remember. However, as noted by Meta spokesperson Emil Vazquez, opting out of these personalized features isnāt offered yet.
6. NVIDIA Welcomes AI Disruption as DeepSeek Emerges š
Chinese AI startup DeepSeekās new R1 model has caused a stir by outperforming top U.S. models like OpenAIās, leading to a notable 17% drop in NVIDIAās stock. NVIDIA, however, welcomed the development, praising DeepSeekās use of āTest Time Scalingā and highlighting that this innovation could increase demand for its GPUs.
The R1 modelās impressive results were achieved at a fraction of the cost typically spent by tech giants, sparking questions about the value of massive AI infrastructure investments by companies like Microsoft and Meta.
Gemini AI Feature: Chat with your PDFs in Google Drive
Thereās a Gemini AI feature inside Google Workspace that lets you talk to your PDFs inside Drive with Gemini.
Lately, Gemini features have been a hit or miss. So is this one any good?
Check out today's AI in 5.
š¦¾How You Can Leverage:
High schoolers might be better at building an AI-native company than seasoned and stuck-in-their-ways vets.
Hereās the story.
A Gartner Fellow walks away from his cushy tech-predicting gig to build an AI nutrition app.
Plot twist?
He staffs it with teenagers who've never written a line of traditional code. And they're crushing it.
Meet Rajesh Kandaswamy, founder of Nourish and former Gartner Researcher. His unconventional bet on Gen Z talent is reshaping what an AI-first company looks like.
No corporate playbooks.
No high-priced software engineers with decades of experience.
No "we've always done it this way." Just pure, unfiltered digital native intuition.
Want the secret sauce for building a true AI-first company?
Rajesh joined Everyday AI today to lay out the playbook for building an AI-first company.
These insights might sting a little if youāre stuck in your ways.
Youāve been warned.
1 ā The 90-Second Gamble š²
Nourish's first version of natural language food tracking took too long to process. Likeā¦. An hour.
Most founders would've ditched it. Not Rajesh.
He bet on computing getting faster ā just like Bezos bet on dial-up improving for early Amazon. Today?
Ten seconds flat.
Computing speed, costs, and connectivity only move in one direction. Smart money bets on that curve.
Try This:
Take your most painful AI process right now. Time it precisely. Now imagine it 100x faster ā what features could you build?
Start architecting those features today. When processing times drop (they will), you'll be ready while everyone else is still planning their first move.
2 ā Why College Grads Canāt Build AI Companies Right Now š
The wildest part of Nourish's story might not be the tech.
High school interns across Dubai, India, China, and the UK are building and shipping code. Zero traditional training. Pure AI-powered development.
Why? These kids never learned the "right way" to build software. Or, antiquated ways that many colleges might be teaching. (Yeah, many traditional colleges are still banning Generative AI instead of infusing it in every subject.)
The young Nourish team didnāt fight the AI ā they flowed with it. No preconceptions means no limitations.
Try This:
Find three teenagers who've never built software the traditional way. Give them an AI suite of tools and a real problem to solve.
Think ā Microsoft Github Copilot, Cursor, Windsurf, Lovable, o1 Pro, etc.
Don't teach them "proper" development. Let them show you how future companies will be built. Your experienced devs might hate it.
That's precisely the point.
3 ā The Death of AI Reliability Fears šµ
Remember when steam engines were death traps?
Now they're in museums. AI's following the same path. But here's what everyone misses:
ChatGPT isn't just tech anymore ā it's a brand. And brands need reliability to survive.
A year ago, asking about calories in an apple got you ten different answers. Today? Rock-solid consistency. The market demands it.
Try This:
Stop tiptoeing around AI hallucination fears. Start building like consistency is inevitable. Map out three features you've been afraid to build because of reliability concerns.
Now design them assuming 99% accuracy.
When the technology catches up ā and it will ā you'll have a six-month head start on the market.
Last thing: Forget everything you know about "digital transformation." This isn't about adding AI to your existing processes.
It's about reimagining your entire operation with AI as the foundation. Anything less is just expensive makeup on an aging face.
Real talk: You're probably doing AI wrong. But that's okay ā even the experts are figuring this out in real time.
The difference? They're not waiting for perfection. They're like Rajesh and crew: building the future with teenage interns, AI and blazing fast agility.
Your competition is already rethinking everything. Are you?
Now This ā¦
Let us know your thoughts!
Vote to see live results
Do you read the entire newsletter? |
Reply