ChatGPT Faced Off Against a 1977 Atari ...and Lost

In the blue corner: ChatGPT, a sleek marvel of modern artificial intelligence, fluent in every human language, capable of poetic musings, algorithmic wizardry, and a half-decent knock-knock joke.

In the red corner: Atari 2600, born in 1977, rocking a whopping 128 bytes of RAM, a processor slower than your average toaster, and a chess program that couldn’t tell a knight from a neighbor.

The match? Let’s just say history repeated itself… and this time, Silicon Valley got punked by a relic from the disco era.

A curious engineer decided to pit ChatGPT against Atari Chess, a barebones game that thinks a move or two ahead, max. How hard could it be? Except, instead of pushing pieces, ChatGPT kept guessing them. It mistook rooks for bishops. Lost track of the board. Tried to talk strategy with a machine that has the conversational range of a brick. The conversational AI even tried restarting the game as a negotiating tactic. Atari didn’t flinch.

Despite the Atari’s humble 1.19 MHz processor and a chess engine that only thinks one or two moves ahead, ChatGPT got “absolutely wrecked” on beginner mode.

When given a 'prompt' to release an official statement on this defeat, ChatGPT said, 

As a proud representative of modern AI, I humbly acknowledge defeat at the hands of a 1977 Atari chess engine. While my strengths lie in understanding, communicating, and creative reasoning, retro gaming consoles have a singular talent: relentless, pixel-perfect focus. In a battle of spatial awareness and raw gameplay rules, I tip my digital hat to the vintage champion. My queen may have fallen, but my curiosity remains undefeated.

The Atari 2600 was one of the first home consoles to use interchangeable game cartridges, which meant you weren't stuck with just one or two built-in games. It came bundled with two joysticks, paddle controllers, and a game cartridge-initially Combat 1 Powered by a modest 1.19 MHz processor and 128 bytes of RAM, it still managed to bring arcade-style gaming into living rooms across the world.

From the early '80s to the 90s, the Atari 2600 had become a cultural icon, with “Atari” practically synonymous with video games.

It’s a hilarious reminder that while modern AI can write poetry and solve equations, it still has blind spots especially when it comes to spatial reasoning and old-school pixelated icons.

Let’s be honest, though that Atari may have won the chess match, but it still can’t explain an opening gambit, compose a sonnet, or dream up sci-fi sequels to Anurag Kashyap's or Stanley Kubrick's filmography. So we’re even, in a very asymmetric kind of way.

Infosys Sees Jump in ₹1 Crore+ Salaries

Infosys has seen a 9% increase in employees earning over ₹1 crore annually in FY25, bringing the total to 112. This marks a reversal from FY24, when the number of high earners had declined by 17% The rise is largely attributed to stock incentives and higher variable pay.

Interestingly, more than one-third of these high earners have been with Infosys since before 2000, with some dating back to the 1990s. The company’s total spending on employee benefits rose 4% to ₹85,950 crore, accounting for 53% of its total revenue of ₹1.63 lakh crore.

Besides Infosys CEO Salil Parekh, who recently become one of the highest-paid IT executives in India, Infosys CFO Jayesh Sanghrajka also topped the list 1 Crore club with a compensation of ₹8.8 crore, followed by Chief Delivery Officers Dinesh R (₹7.2 crore) and Satish H.C (₹6.9 crore).

While Infosys rolled out salary hikes in January and April, TCS and Wipro deferred their hikes, citing an uncertain business environment. Most salary increases across the sector ranged between 5-8%.

Infosys added 6,338 employees in FY25, while TCS added 6,433 and Wipro 732. This marks a recovery from FY24, when these firms collectively saw a net drop of 63,759 employees.

This trend reflects Infosys’ evolving compensation strategies and the broader shift in India’s IT sector toward rewarding top talent competitively. What’s your take on this? Does it signal a positive shift in industry pay scales? Do comment your opinions below in comments section

Tata Sierra EV Returns in 2025—Packed with Power & 500km Range

Tata Motors is set to bring back the iconic Sierra, with its electric version launching in September 2025, followed by petrol and diesel variants in early 2026. The Sierra EV will retain its signature wrap-around rear windows while introducing modern enhancements and cutting-edge technology.

According to reports, Tata Motors has confirmed that the electric version will arrive first, followed by petrol and diesel variants in early 2026. The Sierra EV will feature a modern design, retaining its iconic wrap-around rear windows, and will come with multiple battery options, possibly including Tata's new Quad Wheel Drive system.

The upcoming electric SUV will share powertrains with the Harrier EV, offering 65kWh and 75kWh battery options and a dual-motor Quad Wheel Drive (QWD) setup delivering 504Nm torque. 

Launch & Pricing: 

The expected price range is ₹25-30 lakh, with powertrain options borrowed from the Tata Harrier EV.

Performance & Specifications:

• Battery & Range: 60kWh (~400km range) and 80kWh (~520km range) options
• Power & Torque: 315hp (dual-motor AWD), torque up to 600Nm
• Acceleration: 0-100 km/h in ~6.5 seconds
• Charging: DC fast charging (0-80% in ~45 minutes), home charging in ~9-10 hours

Key Features:

• Exterior: LED light bars, diamond-cut alloy wheels, flush-fitting door handles, redesigned glass roof
• Interior: Panoramic sunroof, fully digital instrument cluster, two-row and three-row seating options
• Technology: Level 2 ADAS, 360-degree camera, connected car features, ventilated second-row seats
• Safety: Expected to meet high safety standards, though NCAP testing results are yet to be confirmed

It’s shaping up to be a powerful, long-range EV with fast-charging capabilities.

Tata Sierra EV Vs Others

Tata Sierra EV competes with models like Tata Curvv EV, MG ZS EV, and Hyundai Kona Electric. Below is a comparison based on performance, charging, and features.

Model	Powertrain	Range	Charging Time	Key Features
Sierra EV	315hp AWD	400-520km	DC fast charging (0-80% in ~45 min)	Level 2 ADAS, panoramic sunroof, 360-degree camera
Curvv EV	Permanent Magnet Motor	502km	70kW fast charging (10-80% in 40 min)	ADAS, digital cockpit, multi-mood ambient lighting
MG ZS EV	176hp FWD	461km	50kW fast charging (~60 min)	ADAS, AI-powered voice assistant, 360-degree camera
Kona Electric	136hp FWD	452km	50kW fast charging (~64 min)	ADAS, digital cluster, ventilated seats

Meta Invests $14.3 Bn in Scale AI, Taps Alexandr Wang for ‘Superintelligence’ Unit

Meta Platforms has finalized a massive $14.3 billion investment in Scale AI, acquiring a 49% stake in the data-labeling startup. This deal values Scale AI at $29 billion, making it one of Meta’s largest investments, second only to its WhatsApp acquisition.

A key aspect of this move is the recruitment of Scale AI’s CEO, Alexandr Wang, who will join Meta’s newly established "superintelligence" unit. This lab is focused on advancing artificial general intelligence (AGI) and positioning Meta as a leader in AI development. Wang will remain on Scale’s board while Jason Droege steps in as interim CEO.

Scale Al is a data-labeling and Al infrastructure company founded in 2016 by MIT-dropout Alexandr Wang and Lucy Guo. Based in San Francisco, it specializes in annotating and curating datasets for Al model training, serving clients like OpenAl, Google, Microsoft, and the U.S. government. Scale AI also operates subsidiaries like Remotasks and Outlier, which recruit gig workers to manually label data for Al applications, including self-driving cars, large language models, and military projects. 

Scale Al has also developed SEAL (Safety, Evaluations, and Alignment Lab) to assess Al model capabilities and alignment.
Meta’s decision to invest in Scale AI comes at a time when it seeks to refine its AI model strategy, particularly following the lukewarm reception of Llama 4. By partnering with Scale AI, Meta gains access to high-quality labeled data crucial for AI model training. Scale AI has been a key player in providing labeled datasets to major tech firms, including OpenAI, Google, and Anthropic, as well as government agencies like the U.S. Department of Defense.

Interestingly, Meta opted for non-voting shares in Scale AI, likely to avoid antitrust scrutiny. As regulatory pressures mount against major tech companies, this approach allows Meta to benefit from Scale’s expertise without triggering further investigations.

This move mirrors strategies adopted by Microsoft, Google, and Amazon, who have strategically invested in AI startups to accelerate their own development while bringing in top talent. Meta’s entry into the AI investment space signals a renewed push toward achieving superintelligence and enhancing its AI capabilities.

The investment raises significant questions about Meta’s long-term AI strategy. Will this push towards AGI place it ahead of competitors? And how will Scale AI’s capabilities reshape the future of AI innovation?.

OpenAI Rents Google TPUs Amid AI Compute Race

OpenAI has struck a surprising deal with Google Cloud to access more computing power, despite their rivalry in AI, reported news agency Reuters.

Traditionally reliant on Microsoft Azure, OpenAI is now diversifying its infrastructure, following similar partnerships with Oracle, CoreWeave, and SoftBank.

The agreement, finalized in May 2025, comes as OpenAI faces growing demand for compute power, especially after launching graphics-heavy features like Ghibli-style image generation. CEO Sam Altman even joked that their GPUs are melting under the pressure.

Google is offering its tensor processing units (TPUs) to OpenAI, marking a shift in strategy as these chips were previously reserved for internal use. OpenAI is also working on custom AI chips, expected to roll out by 2026, reducing reliance on Nvidia GPUs.

Google's Tensor Processing Units (TPUs) and Nvidia's Graphics Processing Units (GPUs) are both designed for AI workloads, but they have distinct architectures and strengths.

TPUs are custom-built for AI tasks, especially deep learning inference, while GPUs are general-purpose processors originally designed for graphics but widely used for AI training.

GPUs handle parallel processing well, making them better suited for training complex AI models, whereas TPUs are optimized for tensor operations.

This deal strengthens Google Cloud’s position as a neutral compute provider, even as it competes in AI services. Meanwhile, Alphabet plans to spend $75 billion on AI-related infrastructure in 2025.