OpenAI raises big money. But can it ever make money? Hashtag Trending for Friday, October 4, 2024

Share post:

Hi, it’s Jim.  One more reminder about CDW Canada Tech Talks. If you’re passionate about technology and innovation, this is the podcast for you.

It’s hosted by KJ Burke, an industry expert. He’ll dive into the latest trends, insights, and strategies shaping the tech landscape in Canada. From hybrid cloud to AI adoption, CDW Canada Tech Talks covers it all. Don’t miss out—visit cdw.ca/techtalks to tune in today. 

And if you want to find out more about K. J. – he’s actually my guest this week on our weekend show. Check it out this weekend. 

HEADLINES

Open AI raises 6.6 billion dollars in a funding round, despite concerns about it’s shift to a for profit model and whether it will ever turn a profit.  NVIDIA drops a bombshell with a new open source model. And Google is back with visual search – will they get it right this time?

Welcome to Hashtag Trending. I’m your host, Jim Love.  Let’s get into it.

Open AI raises 6.6 billion dollars in a funding round

OpenAI’s record-breaking $6.6 billion funding round and the financial hurdles that lie ahead as the company shifts toward a for-profit model.

OpenAI has secured the largest venture capital deal ever, raising $6.6 billion and valuing the company at an astonishing $157 billion. The funding was led by Thrive Capital and included heavyweights like Microsoft, Nvidia, SoftBank, and others. Notably, Apple was absent from the investor list despite previous talks.

This massive investment comes as OpenAI transitions from a nonprofit lab to a product-focused, for-profit enterprise. Investors have the option to withdraw their funding if this shift isn’t completed within two years. OpenAI stated, ‘The new funding will allow us to double down on our leadership in frontier AI research.’

However, the company faces soaring costs. Training new AI models demands immense computing power and energy, with expenses potentially hitting $3 billion this year. Operating services like ChatGPT adds another layer of cost, nearing $4 billion, especially as user numbers have reportedly skyrocketed to 350 million monthly users in June.

Despite expecting $3.7 billion in revenue this year, OpenAI is projected to lose $5 billion, highlighting the financial challenges of leading the AI revolution. The company also faces internal strains; CTO Mira Murati and two top researchers recently departed amid reports of culture clashes between product development and safety teams.

I’m not vouching for the accuracy of this report, but I read a very lengthy and what looked to be thorough analysis of why OpenAI will never make a profit – chief among this was the idea that AI isn’t scalable.  The report predicts that OpenAI will make major increases in how much it charges, but that won’t even be enough.  Every new user adds to the processing costs, with no economy of scale and every power user loses them far more than the 20 to 40 dollars that OpenAI is charging. 

OpenAI is a bad business

With all of that, with recent departures and investors who are going to be looking for returns, OpenAI has a lot of challenges.  And they just got at least two more.

Nvidia drops a bombshell with a new Open Source AI 

Nvidia dropped a bombshell into the AI world when it unveiled NVLM 1.0, an open-source AI model that reports say rivals industry leaders like OpenAI’s GPT-4. Despite these claims, the flagship model, NVLM-D-72B, boasts 72 billion parameters versus GPT4’s 1.8 trillion parameters. But despite the, and I put this in quotes, “compact” size of NVIDIA’s model, it’s reported to excel in both vision and language tasks as well as complex math, in short, some of the key elements that were making OpenAI’s recent model launches real breakthroughs in AI.

Also remarkable is Nvidia’s decision to make the model weights publicly available and to release the training code. This breaks from the norm of keeping advanced AI systems closed, granting researchers and developers unprecedented access to cutting-edge technology.

NVLM-D-72B isn’t just powerful—it’s versatile. It can interpret memes, analyze complex images, and solve mathematical problems step-by-step. The researchers highlighted, ‘Our NVLM-D-1.0-72B demonstrates significant improvements over its text backbone on text-only math and coding benchmarks.’

This is going to send a shockwave through the AI community. One researcher noted, ‘Nvidia just published a 72B model that’s on par with Llama 3.1 405B in math and coding evaluations and also has vision capabilities.’

This could reshape the AI landscape, leveling the playing field for smaller organizations and independent researchers. It’s certainly going to put some real questions about what this does to the commercial AI marketplace. Equally, or maybe more of an issue is that it also raises questions about ethical implications and the future of AI business models.

Google makes search visible.

And if Nvidia’s announcement wasn’t enough, Google drove hard into another area that OpenAI was working on – search.

Starting today, Google is rolling out a new feature that lets you search the web by simply taking a video. If this sounds familiar, Google has launched a couple of similar offerings in the past months, only to have them flop miserably. One that did visual recognition turned out to be an edited version of the real demo video that made the model look much better than it was. And a second more recent launch gave some responses that went viral because of their weirdness – like putting glue on pizza to keep the toppings from sliding off. 

But you can never count Google out. The are back with a new model that promises a lot.

According to Google, you can point your smartphone at something intriguing, asking a question, and instantly receiving answers. For example, you’re at an aquarium, and you wonder why a group of fish swims in unison. Just film them, ask your question, and Google’s AI will analyze the video and provide insights.

Liz Reid, Google’s head of search, says this feature aims to make asking questions about the world around us more effortless. ‘This new capability allows people to ask questions about the world more easily,’ she explains.

Industry analyst Paolo Pescatore believes this is a significant move for Google. ‘We are now seeing AI in everything, and people connect best with visuals,’ he notes. ‘This latest development further demonstrates new ways of bringing content to life with something like search.’

But that’s not all. Google is also enhancing shopping results with more reviews and pricing information and introducing a new tool to identify songs directly from websites or streaming apps, challenging Apple’s Shazam.

We’ll see and hear more about this in the coming days.

And that’s our show for today. 

Thanks to our sponsor, CDW and KJ Burke’s CDW Canada Tech Talks. 

Send me a note at editorial@technewsday.ca 

I’m your host Jim Love, have a Fantastic Friday.

SUBSCRIBE NOW

Related articles

Deep Seek and Open Source AI – Without the Hype: Discussion with Robert Falzon, Head of Engineering, Check Point

DeepSeek AI is shaking up the cybersecurity world—are we prepared for the risks? Join host Jim Love and...

Can Canada Get Its Mojo Back? Interview with Senator Colin Deacon

In this weekend edition of Hashtag Trending, host Jim Love delves into the challenges Canada faces as it...

Deep Seek AI Revolution: Project Synapse on Hashtag Trending for January 25, 2025

Discover how DeepSeek's groundbreaking open-source AI model, R1, is revolutionizing the Artificial Intelligence landscape and redefining global tech...

Project Stargate – 500 Billion Dollars In AI Investment. How Real Is It? Hashtag Trending for Friday, January 24, 2025

ChatGPT Outage Follows On Project Stargate Announcement, Broadcom’s VMWare Lock-In Is Still Angering Customers, And Devin, Reported As...

Become a member

New, Relevant Tech Stories. Our article selection is done by industry professionals. Our writers summarize them to give you the key takeaways