‘Distillation’ Is AI’s New Buzzword—and a Scary One for AI Companies

Summary

WSJ Tech News Briefing: ‘Distillation’ Is AI’s New Buzzword—and a Scary One for AI Companies
Release Date: February 6, 2025
Host: Pierre Bienname, The Wall Street Journal

---

Introduction

On February 6, 2025, The Wall Street Journal’s Tech News Briefing, hosted by Pierre Bienname, delved into two pivotal topics shaping the tech landscape: Amazon’s retreat from its brick-and-mortar ventures and the emerging trend of “distillation” in artificial intelligence (AI) development. This episode provides insightful discussions with industry experts Kate King and Miles Krupa, examining the implications of these developments for businesses and investors alike.

---

Amazon’s Shift Away from Physical Stores

Overview:
Amazon, a titan in online retail, has been gradually scaling back its physical store presence. Despite innovative concepts like Amazon Go, the company has been closing various storefronts, raising questions about the future of its brick-and-mortar strategy.

Key Points:

1. Amazon Go and Customer Reception:
   Amazon Go stores, renowned for their cashless and checkout-free experience, have seen a significant reduction in numbers—from 32 to 16 locations across four states since early 2023. While sales have grown at Whole Foods Market, reaching $5.2 billion in Q3 2024, this pales in comparison to Amazon’s $61.4 billion online sales.

   • Kate King, Real Estate Reporter:
     “I did interview one person who said that it was convenient in the moment to go into the store and walk out without waiting in line. But it didn't necessarily change his life...” (02:09)

2. Challenges with Innovative Retail Concepts:
   Customers appreciate the convenience of Amazon Go but often find traditional payment methods like Apple Pay sufficiently quick, diminishing the perceived benefits of the new technology.

   • King:
     “Paying by credit card is so quick now... so for him it wasn't a huge game changer.” (01:44)

3. Diverse Retail Experiments and Their Decline:
   Amazon has experimented with various store formats, including bookstores, Four Star stores (featuring best-selling online items), and tech-driven fashion stores. However, these ventures have been largely abandoned in favor of focusing on more profitable segments.

   • King:
     “Amazon's tried a bunch of different bricks and mortar concepts... Amazon has closed and kind of abandoned those projects.” (02:12)

4. Focus on Grocery and Technological Adaptations:
   The acquisition of Whole Foods in 2017 for over $13 billion remains a success, with sales and locations expanding. Amazon Fresh, a high-tech grocery store, initially implemented Amazon Go’s “just walk out” technology but later reverted to using shopping carts with scanners after encountering operational challenges.

   • King:
     “Even some of the harshest critics that I interviewed do say that the Amazon Fresh redesign has worked...” (02:44)

5. Licensing and Future Prospects of Just Walk Out Technology:
   Despite scaling back its own Amazon Go stores, Amazon continues to develop and license its checkout-free technology to over 200 third-party retailers, including institutions like colleges, hospitals, and convenience stores in high-traffic areas.

   • King:
     “Amazon licenses this technology to over 200 third party retailers... [it] wasn't a waste of time or money...” (03:53)

Conclusion:
Amazon is recalibrating its physical retail strategy by shuttering less successful ventures while leveraging its innovative technologies through licensing. The focus on Whole Foods and the redesigned Amazon Fresh indicates a strategic pivot towards more sustainable and scalable brick-and-mortar operations.

---

The Rise of Distillation in AI and Its Implications

Overview:
The episode shifts focus to the AI industry, highlighting Deepseek, a Chinese AI company that has garnered attention for its cost-effective and efficient AI model training through a process known as distillation. This technique poses significant challenges for established AI giants and raises concerns among investors.

Key Points:

1. Understanding Distillation:
   Distillation involves using the outputs of one AI model to train another, enabling the creation of smaller, more efficient models that retain much of the original’s capabilities. This method allows smaller companies to develop competitive AI models without the immense resources typically required.

   • Miles Krupa, Tech Reporter:
     “It's basically taking the outputs of one AI model and using that to sort of teach another one how to solve problems and ask questions.” (06:18)

2. Deepseek’s Strategic Use of Distillation:
   Deepseek’s AI model, R1, has been developed by distilling knowledge from established models like those from OpenAI. This approach has enabled Deepseek to produce high-performing AI models at a fraction of the cost, intensifying competition in the AI space.

   • Krupa:
     “Ali Godse... compared distillation to basically getting to ask any question you want of Einstein...” (06:18)

3. Impact on Major AI Players and Investors:
   The effectiveness of distillation challenges the investments of hyperscalers—companies like Google, Meta, and OpenAI—that have heavily invested in developing proprietary AI technologies. The ability to replicate advanced AI capabilities cheaply questions the sustainability of these large-scale investments.

   • Krupa:
     “It just raises the question of what's the point if everything can be replicated so quickly and so cheaply.” (08:33)

4. Cost Reduction and Commodification of AI:
   Distillation accelerates the commodification of AI models, driving down costs and increasing accessibility. This trend is likely to spur further replication and innovation, potentially diminishing the unique advantages held by early AI pioneers.

   • Krupa:
     “Everything is coming down in cost. A lot of people are talking about the commodification of the AI models...” (09:33)

5. Challenges in Enforcing Intellectual Property Rights:
   OpenAI, whose terms of service prohibit using its models to develop competing products, faces difficulties in enforcing these restrictions. Efforts to shut down accounts linked to infringing activities are often temporary solutions in the ongoing battle against misuse.

   • Krupa:
     “It's hard. They said that they've closed down an account they believe was tied to Deepseek... it's a game of whack a mole.” (10:17)

Conclusion:
The advent of distillation presents a paradigm shift in AI development, democratizing access to powerful models but simultaneously undermining the investments of major AI companies. As the technology becomes more accessible and cost-effective, the competitive landscape is poised for significant transformation, raising critical questions about the future of AI innovation and investment.

---

Final Thoughts

This episode of WSJ Tech News Briefing underscores the dynamic and evolving nature of the tech industry. Amazon’s strategic retreat from physical stores, coupled with its continued investment in innovative technologies, reflects broader trends in consumer behavior and retail operations. Simultaneously, the rise of distillation in AI development highlights a potential inflection point in the AI race, where efficiency and cost-effectiveness could redefine competitive advantages and industry standards.

For listeners seeking to stay abreast of these developments, the discussions presented by Kate King and Miles Krupa offer valuable perspectives on the challenges and opportunities facing both retail giants and emerging AI companies.

---

Produced by Julie Chang with supervising producer Kathryn Millsop.

Transcript

[ADP Representative]

ADP knows any big thing, any small thing, any trendy thing, even a trendy thing that everyone knows isn't a great idea, but management just wants us to give it a try for a bit. Can change the world of work. From HR to payroll, ADP designs forward thinking solutions to take on the next anything.

[Pierre Bienname]

Welcome to Tech News briefing. It's Thursday, February 6th. I'm Pierre Bienname for the Wall Street Journal. Amazon may be number one in online retail, but its bricks and mortar stores aren't doing quite as well. We'll talk to real estate reporter Kate King about the company's decision to shut some of its storefronts. And Deepseek, the buzzy artificial intelligence company out of China, may have trained one of its models by having it ask questions of its American counterparts. Tech reporter Miles Krupa says so called distillation has some investors spooked. First, after a decade long experiment with real life stores, Amazon is pulling back. In recent years, the e commerce company has closed all kinds of shops. Its portfolio of Amazon Go convenience stores, where you can just grab an item and walk out and you'll be charged electronically, has shrunk by about half since early 2023-16 stores in four states. Sales at its stores, including Whole Foods Market, have grown annually, topping $5.2 billion in the third quarter of 2024. But that compares with about $61.4 billion at Amazon's online store. Kate King covers real estate for the Wall Street Journal and she joins me now. Kate, Amazon Go stores. They do seem pretty innovative. There's no traditional checkout counter, so there's no line and no one likes a line. They use these cameras and sensors to track your purchases. Are customers just not into it?

[Kate King]

I'm sure some are for sure. I did interview one person who said that it was convenient in the moment to go into the store and walk out without waiting in line. But it didn't necessarily change his life and he said that while maybe he saves a few seconds not having to pull out his credit card, paying by credit card is so quick now, especially with Apple Pay, which allows you to quickly just scan with your phone at the checkout aisle. So for him it wasn't a huge game changer.

[Pierre Bienname]

And it's not just Amazon Go stores, right?

[Kate King]

Amazon's tried a bunch of different bricks and mortar concepts. That includes of course their bookstores, which they started with about 10 years ago in Seattle. Their four star locations which were stores that were stocked with their best selling items from the website. They had their Amazon style Stores, which were basically fashion stores with a lot of technology. And of course they have their grocery portfolio and the other types of bricks and mortar stores, the bookstores, the four star, the fashion stores. Amazon has closed and kind of abandoned those projects. What they're really focusing now on is.

[Pierre Bienname]

Their grocery portfolio, namely Whole Foods, which it owns.

[Kate King]

Yes, Amazon bought Whole foods market in 2017 for over $13 billion. And it's doing really well. Sales are growing, it's expanded the locations of Whole Foods. And Amazon also has launched a concept called Amazon Fresh, which is a high tech grocery store. It sells a lot of mass market items at maybe more affordable prices than you'd find at Whole Foods. And it uses technology. And here Amazon's really experimented and made some major changes. When it first opened the Amazon fre, they had the just walk out technology that you see at Amazon Go convenience stores. But this didn't really work. And so they paused expansion of Amazon Fresh. They removed the just walk out technology and now they have shopping carts that are equipped with scanners. So when you pick up an item off the shelf, you scan it, you put it into the cart. And even some of the harshest critics that I interviewed do say that the Amazon Fresh redesign has worked. But at the same time, it remains to be seen how successful it will be in terms of growth and customer adoption.

[Pierre Bienname]

So is Amazon scrapping its just walkout tech?

[Kate King]

Amazon is still using its just walkout technology. While it might be walking away or paring down its focus on the Amazon Go stores, it's not reducing its emphasis on just walkout technology. Amazon licenses this technology to over 200 third party retailers. That includes colleges and universities, hospitals, those convenience type stores in airports, stadiums. And someone I interviewed who was really critical of Amazon's physical locations said Amazon Go. While it didn't really work as a standalone store, it wasn't a waste of time or money because it allowed Amazon to really research, develop and refine this technology, which it's now selling to other retailers.

[Pierre Bienname]

That was our reporter, Kate King. Coming up, training. Cutting edge AI is massively expensive. But if the next best AI models are super cheap, where does that leave the pioneers and their investors? That's after the break.

[ADP Representative]

ADP imagines a world of work where smart machines become too smart copier. I need 15 copies of this printing, by the way.

[Kate King]

Irregardless.

[ADP Representative]

Not a word, Janet.

[Kate King]

Yeah, I know.

[Pierre Bienname]

Page 6 should be regardless of or.

[Kate King]

Irrespective of, just print them, please.

[Pierre Bienname]

If it were a word, Janet, it would mean without irregard, which is Copier.

[Kate King]

Switch to silent mode.

[ADP Representative]

Let's put a pin in it. Anything can change the world of work. From HR to payroll. ADP helps businesses take on the next anything.

[Pierre Bienname]

The Chinese company Deepseek turned heads last week with a powerful problem solving AI model, R1. And it was a lot cheaper to make than models made by deep pocketed American companies. But could deepseak have piggybacked on other artificial intelligence? That's the idea behind distillation. By drawing on the results of others work, distillation can create a model that's almost as good quickly and more cheaply. OpenAI, the maker of ChatGPT, says it's seen indications that DeepSeq distilled from the AI models that power ChatGPT. DeepSeek didn't respond to emails seeking comment, but the company has said it used distillation on open source AIs released by Meta and Alibaba in the past. Miles Krupa covers tech for the Wall Street Journal and he joins me now to talk about this. Miles, how does distillation work?

[Miles Krupa]

It's basically taking the outputs of one AI model and using that to sort of teach another one how to solve problems and ask questions. And so it's actually a technique that's been around for a while in AI. We've seen examples of distillation in the past from big US AI labs, but DeepSeq certainly is making everybody much more aware of the practice. Now Ali Godse, who runs this company called Databricks that does a lot of AI work, he compared distillation to basically getting to ask any question you want of Einstein and becoming almost as knowledgeable as he is in physics. That's what the process looks like for these AI models.

[Pierre Bienname]

And we should quickly note that News Corp, owner of the Wall Street Journal, has a content licensing partnership with OpenAI. Miles, we reported that OpenAI is probing whether Deepseek used its models to train its chatbot using distillation.

[Miles Krupa]

That's right. OpenAI began as this company that wanted to democratize AI research and make it available for all as part of its mission and over time has gotten a lot more closed in terms of protecting its ip. It stopped sharing as much research, and that's true of a lot of AI labs. But OpenAI in particular is very sensitive to its trade secrets or it's the technology that it spent billions of dollars to create being improperly used in competing AI, as Deepseek appears to have done, according to OpenAI.

[Pierre Bienname]

And how effective has distillation proven to be so far?

[Miles Krupa]

From what we've seen, it appears to be really effective. It's an easy way for smaller developers in particular to recreate some of the capabilities of much larger models in a much more cost efficient manner, and doing it in a way that also produces AI models that themselves are a lot smaller and more cost efficient.

[Pierre Bienname]

So if distillation works so well, what effects could that have on the hyperscalers, the companies that are pouring vast sums into AI and their investors?

[Miles Krupa]

Yeah, it just puts even more pressure on them to stay at the cutting edge to justify the billions of dollars they're spending. You know, it just raises the question of what's the point if everything can be replicated so quickly and so cheaply. I mean, it was a matter of months between when OpenAI announced this model called O1, that was a very powerful model. It's something called reasoning. And then a few months later, Deepsea came out with its own version R1 that was basically just as good. And so when investors are looking for justification for the likes of Google Meta OpenAI spending billions on chips and data centers and salaries for researchers to stay on the cutting edge of AI, it just raises questions about whether all of that is really necessary.

[Pierre Bienname]

And in fact, overall, what could this do in the arms race in terms of changing the costs of AI?

[Miles Krupa]

Overall, everything is coming down in cost. A lot of people are talking about the commodification of the AI models that power things like ChatGPT, that basically all of the sort of secrets are going to become secret no longer. And open source developers like DeepSeek will only sort of increase the pace of replication and innovation, and that will all help to drive down costs in the underlying technology. That's where a lot of people see the puck traveling.

[Pierre Bienname]

OpenAI's terms of service forbid using its AI to develop rival products the way that OpenAI alleges DeepSeek did. But can OpenAI really do anything to enforce that?

[Miles Krupa]

Yeah, it's hard. They said that they've closed down an account they believe was tied to deepseek. So that's sort of the only real mechanism. It's like a game of whack a mole. Right, because if you close down one account, somebody can always open up another one. But speaking to people about this, it sounds like the way companies like OpenAI and Google provide their AI models through what's called an API. It's very hard to prevent up front the sort of misuse that OpenAI is alleging.

[Pierre Bienname]

That was our reporter Miles Krupa. And that's it for tech news. Briefing. Today's show was produced by Julie Chang with supervising producer Kathryn Millsop. I'm Pierre Bienname for the Wall Street Journal. We'll be back this afternoon with TNB Tech Minute. Thanks for listening.

[ADP Representative]

ADP knows any big thing, any small thing, any trendy thing, even a trendy thing that everyone knows isn't a great idea, but management just wants us to give it a try for a bit. Can change the world of work from HR to payroll. ADP designs forward thinking solutions to take on the next anything.