Organised by the magazine Annals of Improbable Research, the annual prize ceremony held at Harvard University celebrates hilarious research that spurs interest in science.

Ig Nobel Prizes awarded for pigeon-guided missiles, butt-breathing pigs and swimming dead trout - ABC News

Improbable

improbable

Is ByteDance’s Lumi an AI Experiment Disguised as an AI Creator Community?

ByteDance's Lumi የ AI ሙከራ እንደ AI ፈጣሪ ማህበረሰብ ተደብቋል?

¿Es Lumi de ByteDance un experimento de IA disfrazado de comunidad de creadores de IA?

Èske Lumi ByteDance a se yon eksperyans AI degize kòm yon kominote kreyatè AI?

Արդյո՞ք ByteDance-ի Lumi-ն AI փորձ է, որը քողարկված է որպես AI Ստեղծողների համայնք:

ByteDance の Lumi は AI クリエイター コミュニティを装った AI 実験なのか?

ByteDance의 Lumi는 AI 크리에이터 커뮤니티로 위장한 AI 실험일까요?

ByteDance’s Lumi ezali Expérience ya AI oyo e déguisé lokola Communauté ya Créateur ya AI?

Lumi ຂອງ ByteDance ເປັນການທົດລອງ AI ປອມຕົວເປັນຊຸມຊົນຜູ້ສ້າງ AI ບໍ?

Adakah Lumi ByteDance merupakan Eksperimen AI yang Menyamar sebagai Komuniti Pencipta AI?

Je, Lumi ya ByteDance ni Jaribio la AI Limefichwa kama Jumuiya ya Watayarishi wa AI?

Оё Lumi byteDance як озмоиши AI аст, ки ҳамчун ҷомеаи Офарандаи AI пинҳон карда шудааст?

Ngaba i-ByteDance's Lumi liLingelo le-AI eliguqulelwe njengoluntu lwabadali be-AI?

NASA’s Post-ISS Plans May Not Add Up: Here's Why

One data scientist recently argued with me that we should stop using the term “hallucination” and just call these outputs what they really are: errors. The premise is simple—calling them errors would set more reasonable expectations for AI’s behavior. While this argument might make sense for someone coming from a statistics or programming background, it fundamentally misunderstands how AI works and why the term “hallucination” is actually more appropriate, albeit imperfect. The Misunderstanding of AI’s Nature From a statistical perspective, an error is a deviation from a known value or standard. It’s easy to quantify, easy to detect, and there’s a clear sense of right and wrong. If a model predicts a 20% probability when the true probability is 30%, we call it an error. We can measure this deviation, adjust the model, and move forward. For a data scientist accustomed to working within this framework, it’s natural to see LLM outputs that are factually incorrect as just another type of error. If an AI model says, “A circle has three sides,” then it’s clearly made a mistake—like a regression model producing a bizarre outlier. The problem with this comparison is that it applies a narrow definition of error to a system that generates language, not discrete values. AI models like GPT-4 do not “make errors” in the traditional sense because they lack a clear objective standard like those found in statistics. Why We Call It “Hallucination” AI language models generate sequences of tokens—words and phrases—based on probabilistic patterns. These patterns are learned from vast amounts of text data, and the models produce the most probable sequence of words given the context. Because of this structure, a language model might generate a phrase that is syntactically correct but semantically flawed. Imagine these two statements: “A circle has three sides.” “The professor said a circle has three sides.” If you evaluate these on a token-by-token basis, both might be highly probable outputs based on their preceding context. The first statement is objectively false, while the second statement could be contextually true if a professor actually made that erroneous claim. The model doesn’t have an internal understanding of geometry or truth—it’s simply generating words that fit together well. That’s why when the output is nonsense, we call it a “hallucination” instead of an error. The model didn’t make a mistake according to its own mechanics; it just produced an improbable yet plausible-sounding string of text. The Inadequacy of Calling It “Error” Labeling such outputs as errors creates the false impression that these models should know better. But knowing requires understanding, and that’s not how AI models are built. They don’t “know” in the way humans do, nor do they validate statements against a factual baseline. There is no inherent “truth” within the architecture of an LLM. The model operates based on patterns and associations, not logical consistency or factual accuracy. When I hear the argument that AI hallucinations are “logical errors” or “input errors,” it reminds me of the Garbage In, Garbage Out (GIGO) adage in traditional programming. If your inputs are flawed or your logical framework is off, the system produces wrong outputs. But that analogy only goes so far with AI because language and context are far more nuanced than a spreadsheet of values or a database query. An LLM doesn’t make logical errors in the programming sense. Instead, it lacks an internal verification process to ensure that what it’s saying is true. Imagine trying to apply the same logic to artistic styles or creativity. If an AI-generated image combines elements of a cat and a cloud, calling it an “error” is inappropriate. The model didn’t err—it produced a plausible but nonsensical creation based on its input data and generative process. Hallucinations and Human Perception This distinction is crucial because calling these outputs “errors” could mislead people into thinking that LLMs are making simple mistakes that can be fixed with better logic or more data. But hallucinations in LLMs are not bugs that can be patched. They are a byproduct of how these models work—of generating text without a built-in sense of truth. The onus is still on the human user to guide the AI’s output by crafting better prompts, setting clearer parameters, and using post-processing tools to verify facts. This is why the term “hallucination,” while imperfect, is closer to the mark. It conveys that the model is not just producing an error but fabricating new content—fabrications that are a natural consequence of its design, not the result of poor logic or faulty inputs. It serves as a reminder that these models are generative, not analytical. They don’t have a grasp of reality and can produce completely fabricated content even when all the inputs seem perfect. Setting Expectations for AI’s Outputs Ultimately, the data scientist’s argument comes from a place of wanting to simplify AI’s behaviors for the end-user. If we call these outputs “errors,” maybe people will understand AI’s limitations better. But in reality, this simplification does a disservice to the complexity of these models. We need to help people understand that AI models do not think, reason, or understand—they predict and generate. Their hallucinations are not “errors” in a mechanical or statistical sense but are intrinsic artifacts of how they operate. Until we develop AI models with built-in mechanisms for validating truth, we will continue to see outputs that are not just factually wrong but sometimes fantastical. And as we refine our language and expectations around AI, we need to keep terms like “hallucination” in place—if only to remind ourselves that these models are not yet capable of true understanding. So, while I appreciate the sentiment behind labeling these outputs as errors, it inadvertently strips away the nuances that make AI hallucinations fundamentally different from the statistical errors many data scientists are familiar with. As we continue to refine our models and our language, it’s worth keeping these distinctions in mind to set clearer expectations for what AI can and cannot do. About Me: 25+ year IT veteran combining data, AI, risk management, strategy, and education. 4x hackathon winner and social impact from data advocate. Currently working to jumpstart the AI workforce in the Philippines. Learn more about me here: https://docligot.com One data scientist recently argued with me that we should stop using the term “hallucination” and just call these outputs what they really are: errors. The premise is simple—calling them errors would set more reasonable expectations for AI’s behavior. While this argument might make sense for someone coming from a statistics or programming background, it fundamentally misunderstands how AI works and why the term “hallucination” is actually more appropriate, albeit imperfect. The Misunderstanding of AI’s Nature From a statistical perspective, an error is a deviation from a known value or standard. It’s easy to quantify, easy to detect, and there’s a clear sense of right and wrong. If a model predicts a 20% probability when the true probability is 30%, we call it an error. We can measure this deviation, adjust the model, and move forward. For a data scientist accustomed to working within this framework, it’s natural to see LLM outputs that are factually incorrect as just another type of error. If an AI model says, “A circle has three sides,” then it’s clearly made a mistake—like a regression model producing a bizarre outlier. The problem with this comparison is that it applies a narrow definition of error to a system that generates language, not discrete values. AI models like GPT-4 do not “make errors” in the traditional sense because they lack a clear objective standard like those found in statistics. Why We Call It “Hallucination” AI language models generate sequences of tokens—words and phrases—based on probabilistic patterns. These patterns are learned from vast amounts of text data, and the models produce the most probable sequence of words given the context. Because of this structure, a language model might generate a phrase that is syntactically correct but semantically flawed. Imagine these two statements: “A circle has three sides.” “The professor said a circle has three sides.” If you evaluate these on a token-by-token basis, both might be highly probable outputs based on their preceding context. The first statement is objectively false, while the second statement could be contextually true if a professor actually made that erroneous claim. The model doesn’t have an internal understanding of geometry or truth—it’s simply generating words that fit together well. That’s why when the output is nonsense, we call it a “hallucination” instead of an error. The model didn’t make a mistake according to its own mechanics; it just produced an improbable yet plausible-sounding string of text. The Inadequacy of Calling It “Error” Labeling such outputs as errors creates the false impression that these models should know better. But knowing requires understanding, and that’s not how AI models are built. They don’t “know” in the way humans do, nor do they validate statements against a factual baseline. There is no inherent “truth” within the architecture of an LLM. The model operates based on patterns and associations, not logical consistency or factual accuracy. When I hear the argument that AI hallucinations are “logical errors” or “input errors,” it reminds me of the Garbage In, Garbage Out (GIGO) adage in traditional programming. If your inputs are flawed or your logical framework is off, the system produces wrong outputs. But that analogy only goes so far with AI because language and context are far more nuanced than a spreadsheet of values or a database query. An LLM doesn’t make logical errors in the programming sense. Instead, it lacks an internal verification process to ensure that what it’s saying is true. Imagine trying to apply the same logic to artistic styles or creativity. If an AI-generated image combines elements of a cat and a cloud, calling it an “error” is inappropriate. The model didn’t err—it produced a plausible but nonsensical creation based on its input data and generative process. Hallucinations and Human Perception This distinction is crucial because calling these outputs “errors” could mislead people into thinking that LLMs are making simple mistakes that can be fixed with better logic or more data. But hallucinations in LLMs are not bugs that can be patched. They are a byproduct of how these models work—of generating text without a built-in sense of truth. The onus is still on the human user to guide the AI’s output by crafting better prompts, setting clearer parameters, and using post-processing tools to verify facts. This is why the term “hallucination,” while imperfect, is closer to the mark. It conveys that the model is not just producing an error but fabricating new content—fabrications that are a natural consequence of its design, not the result of poor logic or faulty inputs. It serves as a reminder that these models are generative, not analytical. They don’t have a grasp of reality and can produce completely fabricated content even when all the inputs seem perfect. Setting Expectations for AI’s Outputs Ultimately, the data scientist’s argument comes from a place of wanting to simplify AI’s behaviors for the end-user. If we call these outputs “errors,” maybe people will understand AI’s limitations better. But in reality, this simplification does a disservice to the complexity of these models. We need to help people understand that AI models do not think, reason, or understand—they predict and generate. Their hallucinations are not “errors” in a mechanical or statistical sense but are intrinsic artifacts of how they operate. Until we develop AI models with built-in mechanisms for validating truth, we will continue to see outputs that are not just factually wrong but sometimes fantastical. And as we refine our language and expectations around AI, we need to keep terms like “hallucination” in place—if only to remind ourselves that these models are not yet capable of true understanding. So, while I appreciate the sentiment behind labeling these outputs as errors, it inadvertently strips away the nuances that make AI hallucinations fundamentally different from the statistical errors many data scientists are familiar with. As we continue to refine our models and our language, it’s worth keeping these distinctions in mind to set clearer expectations for what AI can and cannot do. About Me: 25+ year IT veteran combining data, AI, risk management, strategy, and education. 4x hackathon winner and social impact from data advocate. Currently working to jumpstart the AI workforce in the Philippines. Learn more about me here: https://docligot.com About Me: 25+ year IT veteran combining data, AI, risk management, strategy, and education. 4x hackathon winner and social impact from data advocate. Currently working to jumpstart the AI workforce in the Philippines. Learn more about me here: https://docligot.com

Why Your Data Scientists Will Struggle With AI Hallucinations

The AI Disruptobloat: How Overproduction Dilutes Value but Accelerates Innovation

AI Disruptobloat: Как свръхпроизводството намалява стойността, но ускорява иновациите

La disrupción de la IA: cómo la sobreproducción diluye el valor pero acelera la innovación

The AI Disruptobloat: como a sobreprodución dilúe o valor pero acelera a innovación

The AI Disruptobloat: כיצד ייצור יתר מדלל ערך אך מאיץ חדשנות

AI による破壊的肥大化: 過剰生産は価値を薄め、イノベーションを加速させる

AI Disruptobloat: ການຜະລິດເກີນຂະໜາດເຮັດໃຫ້ຄຸນຄ່າແຕ່ເລັ່ງການປະດິດສ້າງ

AI disruptobloat: kā pārprodukcija mazina vērtību, bet paātrina inovācijas

AI Disruptobloat: Хэт их үйлдвэрлэл хэрхэн үнэ цэнийг бууруулж, харин инновацийг хурдасгадаг вэ?

The AI Disruptobloat: Sida wax-soo-saarka xad-dhaafka ah u milicsado qiimaha laakiin uu dedejiyo hal-abuurka

AI Disruptobloat: การผลิตมากเกินไปทำให้มูลค่าเจือจางลงแต่เร่งให้เกิดนวัตกรรมเร็วขึ้น

AI disruptobloat: qanday qilib ortiqcha ishlab chiqarish qiymatni kamaytiradi, lekin innovatsiyalarni tezlashtiradi

人工智能颠覆性膨胀：生产过剩如何稀释价值但加速创新

Authors:
(1) STUART KAUFFMAN;
(2) ANDREA ROL. Table of Links Abstract and Introduction Part I. A Definition of Life Part II. The first Miracle: The emergence of life is an expected phase transition – TAP and RAF. Part III. The Second Miracle: The evolution of the biosphere is a propagating, non-deducible construction, not an entailed deduction. There is no Law. Evolution is ever-creative Part IV. New Observations and Experiments: Is There Life in the Cosmos? Conclusion and Acknowledgments Figures and References Part IV. New Observations and Experiments: Is There Life in the Cosmos? i. We are discovering ever more exoplanets, (49,50). We seek evidence of life in the atmospheres of these planets, (50). The results discussed here suggest a potent new hope. If we establish that the small molecule collectively autocatalytic sets in all 6700 prokaryotes do reproduce chemically in vitro, we can then seek evidence in the atmospheres of exoplanets of just such small molecule collectively autocatalytic sets. These sets are plausibly the most rudimentary form of life in the universe. Given an identified set of small molecules in an atmosphere, it is calculation to test if that set can function as a small molecule collectively autocatalytic set, as done by Xavier, (11,12). If Yes, we may be seeing life. ii. It is now feasible to create high diversity small molecule libraries. For example, work by Ott running the Miller Urey experiment starting with only four molecular species for a month yields thousands of small molecules, identified by mass spectrometry, (51). We can now ask if in such systems, small molecule collectively autocatalytic sets can emerge. This is a “Go or No Go” experiment. If No, the theory is probably wrong. iii. If Yes, we can begin to envision testable pathways beyond small molecule collectively autocatalytic sets to such sets becoming the metabolism of peptide RNA autocatalytic sets with which they co-evolve as new Kantian Wholes. In turn these might evolve to template replication, and even to genetic coding, (52,53,54). Real experiments are needed. iv. We begin to envision testable pathways from the earliest small molecule Kantian Wholes to the emergence of prokaryotes. From these to the eukaryote may be a long step, but multicellularity arose six or more times. Is the emergence of complex life so very improbable? We do not know. Perhaps not. This paper is available on arxiv under CC BY 4.0 DEED license. Authors: (1) STUART KAUFFMAN; (2) ANDREA ROL. Authors: Authors: (1) STUART KAUFFMAN; (2) ANDREA ROL. Table of Links Abstract and Introduction Abstract and Introduction Part I. A Definition of Life Part I. A Definition of Life Part II. The first Miracle: The emergence of life is an expected phase transition – TAP and RAF. Part II. The first Miracle: The emergence of life is an expected phase transition – TAP and RAF. Part III. The Second Miracle: The evolution of the biosphere is a propagating, non-deducible construction, not an entailed deduction. There is no Law. Evolution is ever-creative Part III. The Second Miracle: The evolution of the biosphere is a propagating, non-deducible construction, not an entailed deduction. There is no Law. Evolution is ever-creative Part IV. New Observations and Experiments: Is There Life in the Cosmos? Part IV. New Observations and Experiments: Is There Life in the Cosmos? Conclusion and Acknowledgments Conclusion and Acknowledgments Figures and References Figures and References Part IV. New Observations and Experiments: Is There Life in the Cosmos? i. We are discovering ever more exoplanets, (49,50). We seek evidence of life in the atmospheres of these planets, (50). The results discussed here suggest a potent new hope. If we establish that the small molecule collectively autocatalytic sets in all 6700 prokaryotes do reproduce chemically in vitro, we can then seek evidence in the atmospheres of exoplanets of just such small molecule collectively autocatalytic sets. These sets are plausibly the most rudimentary form of life in the universe. Given an identified set of small molecules in an atmosphere, it is calculation to test if that set can function as a small molecule collectively autocatalytic set, as done by Xavier, (11,12). If Yes, we may be seeing life. ii. It is now feasible to create high diversity small molecule libraries. For example, work by Ott running the Miller Urey experiment starting with only four molecular species for a month yields thousands of small molecules, identified by mass spectrometry, (51). We can now ask if in such systems, small molecule collectively autocatalytic sets can emerge. This is a “Go or No Go” experiment. If No, the theory is probably wrong. iii. If Yes, we can begin to envision testable pathways beyond small molecule collectively autocatalytic sets to such sets becoming the metabolism of peptide RNA autocatalytic sets with which they co-evolve as new Kantian Wholes. In turn these might evolve to template replication, and even to genetic coding, (52,53,54). Real experiments are needed. iv. We begin to envision testable pathways from the earliest small molecule Kantian Wholes to the emergence of prokaryotes. From these to the eukaryote may be a long step, but multicellularity arose six or more times. Is the emergence of complex life so very improbable? We do not know. Perhaps not. This paper is available on arxiv under CC BY 4.0 DEED license. This paper is available on arxiv under CC BY 4.0 DEED license. available on arxiv

Is There Life in the Cosmos? New Observations and Experiments

Somnia Looks to Community Governance and Open Standards to Shape Future of Virtual Worlds

সোমনিয়া ভার্চুয়াল ওয়ার্ল্ডের ভবিষ্যত গঠনের জন্য কমিউনিটি গভর্নেন্স এবং উন্মুক্ত মানদণ্ডের দিকে নজর দেয়

Somnia setzt auf Community Governance und offene Standards, um die Zukunft virtueller Welten zu gestalten

Somnia busca la gobernanza comunitaria y los estándares abiertos para dar forma al futuro de los mundos virtuales

Somnia se tourne vers la gouvernance communautaire et les normes ouvertes pour façonner l’avenir des mondes virtuels

सोमनिया आभासी दुनिया के भविष्य को आकार देने के लिए सामुदायिक प्रशासन और खुले मानकों की ओर देख रहा है

Somnia はコミュニティ ガバナンスとオープン スタンダードに着目し、仮想世界の未来を形作る

Somnia는 가상 세계의 미래를 형성하기 위해 커뮤니티 거버넌스와 개방형 표준을 모색합니다.

Somnia busca governança comunitária e padrões abertos para moldar o futuro dos mundos virtuais

Somnia надеется на управление сообществом и открытые стандарты для формирования будущего виртуальных миров

Somnia, Sanal Dünyaların Geleceğini Şekillendirmek İçin Topluluk Yönetişimine ve Açık Standartlara Bakıyor

Somnia hướng tới quản trị cộng đồng và các tiêu chuẩn mở để định hình tương lai của thế giới ảo

Somnia 期待社区治理和开放标准塑造虚拟世界的未来

From improbable to inevitable. Nvidia’s emergence as the frontrunner to become the world’s first $4 trillion stock has been sustained by the generative AI boom. With no signs of slowing, NVDA’s potential appears endless. The third quarter of 2024 began with the same level of momentum on display for Nvidia as KeyBanc analyst John Vinh raised his target price on the stock to $180. This boost of 33% from current levels would see the stock surpass the $4 trillion market capitalization mark and become the first company to do so. These lofty expectations encapsulate the meteoric rise of the semiconductor firm. Just two years ago, NVDA was battling a sustained Wall Street decline amid tech stock sell-offs. The prospect of becoming a $1 trillion stock seemed like a pipedream, let alone mounting a challenge to become the world’s most valuable stock. The generative AI boom changed everything. Bloomberg estimates suggest that the generative AI market alone will grow at a CAGR of 42% over the next 10 years into a $1.3 trillion market by 2032. Nvidia’s status as a market leader in providing AI chipsets ensures that the company will be a driving force in this growth. Strength in Fundamentals Although the long-term sustainability of the generative AI frenzy has been called into question, there’s little doubt that Nvidia has showcased tangible fundamentals to support its impressive Wall Street performance in recent months. In May, the semiconductor giant reported revenues that tripled to $26.04 billion year over year, weighing in well above analyst forecasts. Additionally, Nvidia saw net income rise to $14.88 billion, comfortably up from the $2.04 billion reported a year prior. Nvidia’s buoyant outlook is further supported by forecasts for Q2 2024 revenue reaching $28 billion, plus or minus two percent. This would also fall way above analyst expectations. Driving Force in Generative AI The brilliance of Nvidia’s position as a market leader in providing the computer chips that will power the generative AI boom has drawn some impressive comparisons. According to Jim Keller, a former AMD and Tesla engineer and current CEO of AI firm Tenstorrent, Nvidia is becoming the IBM of the AI era. Given the sheer scale of the ongoing AI boom, Nvidia could surpass IBM in terms of performance and make an unprecedented claim to become the cornerstone of generative AI. This claim could be sustained by the launch of Nvidia’s Blackwell computing processors, which were unveiled in spring 2024. The new products will replace the company’s popular H100 chips and facilitate further revenue generation. The arrival of Blackwell processors comes at a time when an increasing number of firms are seeking ways to incorporate generative AI into their operations. This could help drive sales even higher for Nvidia. However, Nvidia isn’t simply driving the future of AI hardware for enterprises. The firm announced in its June 2024 keynote address plans to integrate its AI services into the world of video games through its ACE-powered Convert Protocol to create fresh on-demand interactions between games and gamers. Given that Microsoft opted to acquire Activision Blizzard for $68.7 billion on the belief that gaming is the future of entertainment and the metaverse, Nvidia’s strides in the landscape could drive its growth further into the future. Is the GenAI Boom Sustainable? Although it appears clear that generative AI will continue its emergence as a disruptive force throughout a vast array of industries, Nvidia’s position as the king of GenAI isn’t necessarily assured. Competition is growing, and more budget-focused firms like Taiwan Semiconductor Manufacturing Company and Advanced Micro Devices are names that could position themselves as cheaper alternatives as the technology continues its emergence. Likewise, many of the tech giants that are currently among Nvidia’s list of customers are intent on taking the production of AI chips in-house. These factors could all eat into NVDA’s dominance. “Investors should exercise caution and conduct thorough research before investing in any field, especially one as dynamic as GenAI,” warns Maxim Manturov, head of investment research at Freedom24. “Success in GenAI depends not only on cutting-edge technology but also on a robust data infrastructure, a skilled workforce, and strategic planning.” While Nvidia’s position as a generative AI market leader appears solid today, there will certainly be challenges to its dominance in the years to come. Should a rival begin generating cheaper AI chipsets, the company’s stock could tumble. Thinking Two Moves Ahead The most important factor in determining Nvidia’s path to $4 trillion is the fact that the company is never content with resting on its laurels. With Nvidia investing billions of dollars into developing leading generative AI computer chips, we’re looking at a company that’s seeking to stay two moves ahead of its rivals. As the market grows, new competitors are likely to emerge, but as one of the world’s largest companies, it’s unlikely that they would have the firepower to go toe-to-toe with Nvidia. With this in mind, Nvidia’s future is in the firm’s hands. If $4 trillion is inevitable, then the likelihood is that CEO Jensen Huang already has $5 trillion and beyond in his sights. From improbable to inevitable. Nvidia’s emergence as the frontrunner to become the world’s first $4 trillion stock has been sustained by the generative AI boom. With no signs of slowing, NVDA’s potential appears endless. The third quarter of 2024 began with the same level of momentum on display for Nvidia as KeyBanc analyst John Vinh raised his target price on the stock to $180 . This boost of 33% from current levels would see the stock surpass the $4 trillion market capitalization mark and become the first company to do so. target price on the stock to $180 target price on the stock to $180 These lofty expectations encapsulate the meteoric rise of the semiconductor firm. Just two years ago, NVDA was battling a sustained Wall Street decline amid tech stock sell-offs. The prospect of becoming a $1 trillion stock seemed like a pipedream, let alone mounting a challenge to become the world’s most valuable stock. The generative AI boom changed everything. Bloomberg estimates suggest that the generative AI market alone will grow at a CAGR of 42% over the next 10 years into a $1.3 trillion market by 2032. Nvidia’s status as a market leader in providing AI chipsets ensures that the company will be a driving force in this growth. CAGR of 42% over the next 10 years CAGR of 42% over the next 10 years Strength in Fundamentals Although the long-term sustainability of the generative AI frenzy has been called into question, there’s little doubt that Nvidia has showcased tangible fundamentals to support its impressive Wall Street performance in recent months. In May, the semiconductor giant reported revenues that tripled to $26.04 billion year over year , weighing in well above analyst forecasts. tripled to $26.04 billion year over year tripled to $26.04 billion year over year Additionally, Nvidia saw net income rise to $14.88 billion, comfortably up from the $2.04 billion reported a year prior. Nvidia’s buoyant outlook is further supported by forecasts for Q2 2024 revenue reaching $28 billion, plus or minus two percent. This would also fall way above analyst expectations. Driving Force in Generative AI The brilliance of Nvidia’s position as a market leader in providing the computer chips that will power the generative AI boom has drawn some impressive comparisons. According to Jim Keller, a former AMD and Tesla engineer and current CEO of AI firm Tenstorrent, Nvidia is becoming the IBM of the AI era . becoming the IBM of the AI era becoming the IBM of the AI era Given the sheer scale of the ongoing AI boom, Nvidia could surpass IBM in terms of performance and make an unprecedented claim to become the cornerstone of generative AI. This claim could be sustained by the launch of Nvidia’s Blackwell computing processors , which were unveiled in spring 2024. The new products will replace the company’s popular H100 chips and facilitate further revenue generation. launch of Nvidia’s Blackwell computing processors launch of Nvidia’s Blackwell computing processors The arrival of Blackwell processors comes at a time when an increasing number of firms are seeking ways to incorporate generative AI into their operations. This could help drive sales even higher for Nvidia. However, Nvidia isn’t simply driving the future of AI hardware for enterprises. The firm announced in its June 2024 keynote address plans to integrate its AI services into the world of video games through its ACE-powered Convert Protocol to create fresh on-demand interactions between games and gamers. fresh on-demand interactions fresh on-demand interactions Given that Microsoft opted to acquire Activision Blizzard for $68.7 billion on the belief that gaming is the future of entertainment and the metaverse, Nvidia’s strides in the landscape could drive its growth further into the future. future of entertainment future of entertainment Is the GenAI Boom Sustainable? Although it appears clear that generative AI will continue its emergence as a disruptive force throughout a vast array of industries, Nvidia’s position as the king of GenAI isn’t necessarily assured. Competition is growing, and more budget-focused firms like Taiwan Semiconductor Manufacturing Company and Advanced Micro Devices are names that could position themselves as cheaper alternatives as the technology continues its emergence. Likewise, many of the tech giants that are currently among Nvidia’s list of customers are intent on taking the production of AI chips in-house . These factors could all eat into NVDA’s dominance. taking the production of AI chips in-house taking the production of AI chips in-house “Investors should exercise caution and conduct thorough research before investing in any field, especially one as dynamic as GenAI,” warns Maxim Manturov, head of investment research at Freedom24. “Success in GenAI depends not only on cutting-edge technology but also on a robust data infrastructure, a skilled workforce, and strategic planning.” While Nvidia’s position as a generative AI market leader appears solid today, there will certainly be challenges to its dominance in the years to come. Should a rival begin generating cheaper AI chipsets, the company’s stock could tumble. Thinking Two Moves Ahead The most important factor in determining Nvidia’s path to $4 trillion is the fact that the company is never content with resting on its laurels. With Nvidia investing billions of dollars into developing leading generative AI computer chips, we’re looking at a company that’s seeking to stay two moves ahead of its rivals. As the market grows, new competitors are likely to emerge, but as one of the world’s largest companies, it’s unlikely that they would have the firepower to go toe-to-toe with Nvidia. With this in mind, Nvidia’s future is in the firm’s hands. If $4 trillion is inevitable, then the likelihood is that CEO Jensen Huang already has $5 trillion and beyond in his sights.

$4 Trillion Appears Inevitable as Nvidia Remains the Star of the Generative AI Boom

The Hidden Hand: Is Paper Bitcoin Suppressing BTC Prices?

লুকানো হাত: কাগজের বিটকয়েন কি বিটিসি দামকে দমন করছে?

Die verborgene Hand: Drückt Papier-Bitcoin den BTC-Preis?

La mano oculta: ¿El Bitcoin de papel está suprimiendo los precios de BTC?

La main cachée : le papier Bitcoin supprime-t-il les prix du BTC ?

छिपा हुआ हाथ: क्या कागजी बिटकॉइन बीटीसी की कीमतों को दबा रहा है?

隠れた手：紙のビットコインは BTC の価格を抑制しているのか？

숨겨진 손: 종이 비트코인이 BTC 가격을 억제하고 있나요?

A mão oculta: o Bitcoin de papel está suprimindo os preços do BTC?

Скрытая рука: бумажный биткойн подавляет цены на BTC?

Gizli El: Kağıt Bitcoin BTC Fiyatlarını Baskılıyor mu?

Bàn tay ẩn giấu: Bitcoin giấy có đang cản trở giá BTC không?

隐藏的手：纸质比特币正在抑制 BTC 价格吗？

Artificial Intelligence has been advancing at an unprecedented rate, causing fear about whether AI will dominate various aspects of our lives in the near future and “take our jobs.” In this article, by leveraging Moore's Law and other technological growth principles, I’m going to make an educated estimation about the state of AI in the coming years. Let’s explore these predictions and explain why AI may not completely take over within the next five years, and what the timeline might look like for AI to reach a truly advanced state. The Current State of AI AI technology has made significant strides in recent years. From natural language processing to computer vision, AI applications are becoming increasingly sophisticated, whether that’s the recent GPT-4o model, Google’s Gemini, or any one of the models in the race to become the most capable AI model. There are lots of people using these models in positive ways to unlock new methodologies in research and education, but there are also lots of people using these technologies for evil such as deepfakes. However, despite these advancements, there are still considerable challenges and limitations that need to be addressed. Current AI systems need massive amounts of data, computational power, and energy to function effectively. Moore's Law and Its Implications for AI Moore's Law, the observation that the number of transistors on a microchip doubles approximately every two years, has been a reliable predictor of technological progress for decades. More generally, Moore’s Law points to technology getting twice as accessible, cheaper, and efficient every two years. According to Moore's Law, we can expect computational power to continue increasing exponentially. However, there are physical and economic limits to this growth that may impact the pace of AI development. Estimating AI's Growth Using Known Laws By extrapolating from Moore's Law and considering other technological growth trends, we can make predictions about AI's capabilities in the coming years. Computational Power If we consider top-of-the-notch GPUs today, let’s say NVIDIA A100, it provides approximately 312 teraflops (TFLOPS) of performance. Moore's Law suggests a doubling of transistors every two years, which implies a potential 8-fold increase in computational power over five years (2^2.5 ≈ 5.66). Future GPU Performance = 312 TFLOPS × 5.66 ≈ 1766 TFLOPS However, the physical limitations of silicon and the economic costs of producing ever-smaller transistors could slow this growth. The increase might realistically be closer to 4 times due to these constraints: Realistic Future GPU Performance=312 TFLOPS×4≈1248 TFLOPS Data Availability and Quality According to projections by IDC (International Data Corporation), the size of the global datasphere will increase from 64.2 zettabytes in 2020 to 175 zettabytes by 2025. It is essential for the training of increasingly complex AI models that data is growing exponentially. But there are still a lot of obstacles to overcome, including concerns about data privacy, data quality, and effectively labeling and processing this data.The availability of high-quality, labeled data is critical, even with an increase in data output. Preprocessing and data labeling are obstacles that will not scale linearly with data increase, according to current trends. Energy Consumption AI training models like GPT-3 require substantial energy. GPT-3 used approximately 1,287 MWh to train. GPT-4, developed approximately 3 years after its predecessor required over 50 gigawatt-hours, so around 50 times the amount it took to train GPT-3. Assuming AI models become more complex and require more data and computational power, energy consumption could grow significantly. If we estimate a similar 50x increase or a 5000% increase in complexity to around 2500 gigawatt-hours for the next model that could be released in the next 5 years, this would require large improvements in energy production and efficiency. Given current energy efficiency improvements of around 10% per year (source: Lawrence Livermore National Laboratory), this only translates to approximately a 60% increase in energy efficiencies over a 5-year period if the energy production technologies don’t improve significantly. The Need for Increased Complexity in AI Models Current AI models struggle with many major issues today. Lack of Up-to-date Information: Current models are trained on static datasets and do not have real-time access to the internet for the most up-to-date information. Hallucinations: AI models sometimes generate information that is plausible-sounding but incorrect or nonsensical. Understanding Context and Nuance: Handling nuanced human language, context, and intent remains a challenge for AI. To address these issues, future AI models will need to incorporate: Real-Time Data Integration: Access to and processing of real-time data from the internet to provide up-to-date responses. Improved Model Architectures: Enhancements in model architectures to reduce hallucinations and improve the reliability of generated content. Contextual Understanding: Advanced natural language processing techniques to better understand and maintain context over long conversations. Increased Model Complexity Solving these problems will significantly increase the complexity of AI models. For example: Real-Time Data Integration: Requires constant data updates and real-time processing capabilities.
Improved Model Architectures: Likely to involve larger and more complex models with more parameters.
Contextual Understanding: Demands more sophisticated training algorithms and potentially longer training times. Let's assume each of these improvements independently doubles the complexity of the AI model. If we start with the current complexity (C), integrating these advancements might lead to: Future Complexity=C×2(real-time data)×2(improved architectures)×2(contextual understanding)=C×2^3 =C×8 Let’s estimate a 5x increase in complexity as a more conservative yet substantial figure. Data Quality and Ethics AI systems depend heavily on high-quality, unbiased data for training and decision-making. While the quantity of data is increasing rapidly, the quality and ethical considerations surrounding this data are equally crucial. Data Generation and Processing With the increase from 64.2 to 175 zettabytes, data generation will likely outpace the current capabilities of processing and labeling. Ensuring data quality and addressing biases require significant resources. Let's consider the human labor and computational resources needed for data labeling. Assuming 1 hour of human labor is needed to label 1 GB of data, labeling 175 zettabytes would require: 175 ZB×1 hour/GB×10^9  GB/ZB =175×10^9  hours This is not feasible with current resources, indicating a substantial gap in data quality and ethical processing. Conclusion: When Will AI Be in a Truly Advanced State In the Next, Say 5 Years? The idea that AI will "take over" in the next five years seems improbable, even if it will surely continue to advance and become more deeply integrated into many industries. There are several obstacles to overcome, including issues with energy usage, data quality, and processing power limitations. AI will not be able to evolve as fast without significant advancements in energy efficiency, data ethics, and technology.It seems more reasonable to anticipate that AI will become truly advanced within the next ten to fifteen years, based on existing patterns and reasonable development estimates. This prolonged period enables the development and successful implementation of essential ethical, technological, and resource-based breakthroughs. Sources and References: Moore, G. E. (1965). Cramming more components onto integrated circuits. Electronics, 38(8), 114-117.
Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.
Markoff, J. (2015). Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots. HarperCollins.
IDC. (2020). Global DataSphere Forecast 2021-2025.
Lawrence Livermore National Laboratory. (2021). Energy Efficiency Trends in Computing. Artificial Intelligence has been advancing at an unprecedented rate, causing fear about whether AI will dominate various aspects of our lives in the near future and “take our jobs.” In this article, by leveraging Moore's Law and other technological growth principles, I’m going to make an educated estimation about the state of AI in the coming years. Let’s explore these predictions and explain why AI may not completely take over within the next five years, and what the timeline might look like for AI to reach a truly advanced state. The Current State of AI AI technology has made significant strides in recent years. From natural language processing to computer vision, AI applications are becoming increasingly sophisticated, whether that’s the recent GPT-4o model, Google’s Gemini, or any one of the models in the race to become the most capable AI model. There are lots of people using these models in positive ways to unlock new methodologies in research and education, but there are also lots of people using these technologies for evil such as deepfakes. However, despite these advancements, there are still considerable challenges and limitations that need to be addressed. Current AI systems need massive amounts of data, computational power, and energy to function effectively. Moore's Law and Its Implications for AI Moore's Law , the observation that the number of transistors on a microchip doubles approximately every two years, has been a reliable predictor of technological progress for decades. More generally, Moore’s Law points to technology getting twice as accessible, cheaper, and efficient every two years. According to Moore's Law, we can expect computational power to continue increasing exponentially. Moore's Law However, there are physical and economic limits to this growth that may impact the pace of AI development. Estimating AI's Growth Using Known Laws By extrapolating from Moore's Law and considering other technological growth trends, we can make predictions about AI's capabilities in the coming years. Computational Power If we consider top-of-the-notch GPUs today, let’s say NVIDIA A100, it provides approximately 312 teraflops (TFLOPS) of performance. Moore's Law suggests a doubling of transistors every two years, which implies a potential 8-fold increase in computational power over five years (2^2.5 ≈ 5.66). Future GPU Performance = 312 TFLOPS × 5.66 ≈ 1766 TFLOPS However, the physical limitations of silicon and the economic costs of producing ever-smaller transistors could slow this growth. The increase might realistically be closer to 4 times due to these constraints: Realistic Future GPU Performance=312 TFLOPS×4≈1248 TFLOPS Data Availability and Quality According to projections by IDC (International Data Corporation), the size of the global datasphere will increase from 64.2 zettabytes in 2020 to 175 zettabytes by 2025. It is essential for the training of increasingly complex AI models that data is growing exponentially. But there are still a lot of obstacles to overcome, including concerns about data privacy, data quality, and effectively labeling and processing this data. The availability of high-quality, labeled data is critical, even with an increase in data output. Preprocessing and data labeling are obstacles that will not scale linearly with data increase, according to current trends. Energy Consumption AI training models like GPT-3 require substantial energy. GPT-3 used approximately 1,287 MWh to train. GPT-4, developed approximately 3 years after its predecessor required over 50 gigawatt-hours , so around 50 times the amount it took to train GPT-3. Assuming AI models become more complex and require more data and computational power, energy consumption could grow significantly. 50 gigawatt-hours If we estimate a similar 50x increase or a 5000% increase in complexity to around 2500 gigawatt-hours for the next model that could be released in the next 5 years, this would require large improvements in energy production and efficiency. 2500 gigawatt-hours Given current energy efficiency improvements of around 10% per year (source: Lawrence Livermore National Laboratory), this only translates to approximately a 60% increase in energy efficiencies over a 5-year period if the energy production technologies don’t improve significantly. The Need for Increased Complexity in AI Models Current AI models struggle with many major issues today. Lack of Up-to-date Information: Current models are trained on static datasets and do not have real-time access to the internet for the most up-to-date information. Lack of Up-to-date Information: Hallucinations: AI models sometimes generate information that is plausible-sounding but incorrect or nonsensical. Hallucinations: Understanding Context and Nuance: Handling nuanced human language, context, and intent remains a challenge for AI. Understanding Context and Nuance: To address these issues, future AI models will need to incorporate: Real-Time Data Integration: Access to and processing of real-time data from the internet to provide up-to-date responses. Real-Time Data Integration: Access to and processing of real-time data from the internet to provide up-to-date responses. Real-Time Data Integration: Improved Model Architectures: Enhancements in model architectures to reduce hallucinations and improve the reliability of generated content. Improved Model Architectures: Enhancements in model architectures to reduce hallucinations and improve the reliability of generated content. Improved Model Architectures: Contextual Understanding: Advanced natural language processing techniques to better understand and maintain context over long conversations. Contextual Understanding: Advanced natural language processing techniques to better understand and maintain context over long conversations. Contextual Understanding: Increased Model Complexity Increased Model Complexity Solving these problems will significantly increase the complexity of AI models. For example: Real-Time Data Integration: Requires constant data updates and real-time processing capabilities. Improved Model Architectures: Likely to involve larger and more complex models with more parameters. Contextual Understanding: Demands more sophisticated training algorithms and potentially longer training times. Real-Time Data Integration: Requires constant data updates and real-time processing capabilities. Real-Time Data Integration: Improved Model Architectures: Likely to involve larger and more complex models with more parameters. Improved Model Architectures: Contextual Understanding: Demands more sophisticated training algorithms and potentially longer training times. Contextual Understanding: Let's assume each of these improvements independently doubles the complexity of the AI model. If we start with the current complexity (C), integrating these advancements might lead to: Future Complexity=C×2(real-time data)×2(improved architectures)×2(contextual understanding)=C×2^3 =C×8 Future Complexity=C×2(real-time data)×2(improved architectures)×2(contextual understanding)=C×2^3 =C×8 Let’s estimate a 5x increase in complexity as a more conservative yet substantial figure. Let’s estimate a 5x increase in complexity as a more conservative yet substantial figure. Data Quality and Ethics AI systems depend heavily on high-quality, unbiased data for training and decision-making. While the quantity of data is increasing rapidly, the quality and ethical considerations surrounding this data are equally crucial. Data Generation and Processing With the increase from 64.2 to 175 zettabytes, data generation will likely outpace the current capabilities of processing and labeling. Ensuring data quality and addressing biases require significant resources. Let's consider the human labor and computational resources needed for data labeling. Assuming 1 hour of human labor is needed to label 1 GB of data, labeling 175 zettabytes would require: 175 ZB×1 hour/GB×10^9  GB/ZB =175×10^9  hours 175 ZB×1 hour/GB×10^9  GB/ZB = 175×10^9  hours 175×10^9  hours This is not feasible with current resources, indicating a substantial gap in data quality and ethical processing. Conclusion: When Will AI Be in a Truly Advanced State In the Next, Say 5 Years? The idea that AI will "take over" in the next five years seems improbable, even if it will surely continue to advance and become more deeply integrated into many industries. There are several obstacles to overcome, including issues with energy usage, data quality, and processing power limitations. AI will not be able to evolve as fast without significant advancements in energy efficiency, data ethics, and technology. It seems more reasonable to anticipate that AI will become truly advanced within the next ten to fifteen years, based on existing patterns and reasonable development estimates. This prolonged period enables the development and successful implementation of essential ethical, technological, and resource-based breakthroughs. Sources and References : Sources and References Moore, G. E. (1965). Cramming more components onto integrated circuits. Electronics, 38(8), 114-117. Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press. Markoff, J. (2015). Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots. HarperCollins. IDC. (2020). Global DataSphere Forecast 2021-2025. Lawrence Livermore National Laboratory. (2021). Energy Efficiency Trends in Computing. Moore, G. E. (1965). Cramming more components onto integrated circuits. Electronics , 38(8), 114-117. Electronics Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning . MIT Press. Deep Learning Markoff, J. (2015). Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots. HarperCollins . HarperCollins IDC. (2020). Global DataSphere Forecast 2021-2025. Lawrence Livermore National Laboratory. (2021). Energy Efficiency Trends in Computing.

Predicting the Future of AI With Moore's Law: When Will It Really Take Over?

I Asked Gemini 1.5 Pro About AGI Twice. This Was Its Shocking Answer.

Stories, News & Mentions about IMPROBABLE

Improbable’s Stories on HackerNoon

Improbable's News & Mentions