Connect with us

Accounting

DeepSeek’s overnight fame strains its systems, draws attacks

Published

on

Talk of an artificial-intelligence upstart in China behind a formidable ChatGPT rival had been building for days. 

At the World Economic Forum in Davos last week, some mentioned Hangzhou-based DeepSeek and its recently released R1 model as a prime reason for countries such as the U.S. to be doubling down on AI advancements. On tech chat boards, engineers had begun comparing its programming performance to leading models from the likes of OpenAI and Microsoft Corp. Its product quietly rose through the ranks of top performers on a UC Berkeley-affiliated AI leaderboard. 

Then, within the past 36 hours, interest in the startup exploded. Silicon Valley heavyweights including investor Marc Andreessen and AI godfather and chief Meta Platforms Inc. scientist Yann LeCun began piling into the conversation, with Andreessen calling DeepSeek’s model “one of the most amazing and impressive breakthroughs” he’s ever seen.  

By the end of the weekend, DeepSeek’s AI assistant had rocketed to the top of Apple Inc.’s iPhone download charts and ranked among the top downloads on Google’s Play Store, straining the startup’s systems so much that the service went down for more than an hour. The company was eventually forced to limit signups to those with mainland China telephone numbers — but claimed the move was the result of “large-scale malicious attacks” on its services.

The fallout from the seemingly overnight surge in interest around DeepSeek was swift, and severe: The company’s AI model, which it claims to have developed at a fraction of the cost of rivals without meaningfully sacrificing performance, drove a nearly $1 trillion rout in U.S. and European technology stocks as investors questioned the spending plans of some of America’s biggest companies. The share plunge in AI chipmaker Nvidia Corp. alone erased roughly $279 million in market value, the biggest wipeout in U.S. stock-market history. 

By Monday afternoon, it was clear the overwhelming interest in DeepSeek’s services was taking a toll on the company’s system. “Currently, only registration with a mainland China mobile phone number is supported,” the startup said on its status page. DeepSeek did not specify whether the signup curbs are temporary or how long they will last.

It was the company’s longest major outage since it started reporting its status. Unlike some rivals, DeepSeek’s assistant shows its work and reasoning as it addresses a user’s written query or prompt. Reviews on Apple’s app store and on Alphabet Inc.’s Android Play Store praised that transparency.

Founded by quant fund chief Liang Wenfeng, DeepSeek’s open-sourced AI model is spurring a rethink of the billions of dollars that companies have been spending to stay ahead in the AI race. 

“While it remains to be seen if DeepSeek will prove to be a viable, cheaper alternative in the long term, initial worries are centered on whether U.S. tech giants’ pricing power is being threatened and if their massive AI spending needs re-evaluation,” said Jun Rong Yeap of IG Asia.

Like all other Chinese-made AI models, DeepSeek self-censors on topics deemed politically sensitive in China. Unlike ChatGPT, DeepSeek deflects questions about Tiananmen Square, President Xi Jinping or the possibility of China invading Taiwan. That may prove jarring to international users, who may not have come into direct contact with Chinese chatbots earlier. 

The initial success provides a counterpoint to expectations that the most advanced AI will require increasing amounts of computing power and energy —- an assumption that has driven shares in Nvidia and its suppliers to all-time highs.

The exact cost of development and energy consumption of DeepSeek are not fully documented, but the startup has presented figures that suggest its cost was only a fraction of OpenAI’s latest models. That a small and efficient AI model emerged from China, which has been subject to escalating U.S. trade sanctions on advanced Nvidia chips, is also challenging the effectiveness of such measures.

“The U.S. is great at research and innovation and especially breakthrough, but China is better at engineering,” computer scientist Kai-Fu Lee said earlier this month at the Asian Financial Forum in Hong Kong. “In this day and age, when you have limited compute power and money, you learn how to build things very efficiently.”

For its part, Nvidia — the biggest provider of chips used to train AI software — described DeepSeek’s new model as an “excellent AI advancement” that fully complies with the U.S. government’s restrictions on technology exports. The startup’s work “illustrates how new models can be created” using a technique known as test time scaling, the company said. 

Nvidia’s statement appeared to dismiss some analysts’ and experts’ suspicions that the Chinese startup couldn’t have made the breakthrough it has claimed. The company also pointed out that inference, the work of actually running AI models and using it to process data and make predictions, nonetheless requires a lot of its products.

“Inference requires significant numbers of Nvidia GPUs and high-performance networking,” the company said.

Continue Reading

Accounting

ISSB ISSB proposes amendments to ease application of standards

Published

on

The International Sustainability Standards Board today published an exposure draft proposing amendments to IFRS S2, “Climate-related Disclosures.”

The proposed amendments provide reliefs to ease the application of requirements related to greenhouse gas emissions disclosures. It aims to make it easier for companies to apply the standards while retaining decision-useful information for investors. The exposure draft will be open for comment for 60 days and close on June 27. The amendments are slated to be finalized by the end of the year, subject to stakeholder feedback. 

“It is the role of a responsible standard-setter to listen to market feedback from the earliest implementation stages, and to support preparers in the application of our Standards,” ISSB vice-chair Sue Lloyd said in a statement. “As a market-focused standard-setter, we have taken steps to respond in a timely manner by proposing targeted amendments helping preparers where possible, without causing too much disruption and ensuring that our Standards continue to enable the provision of decision-useful information to investors.”

ifrs-foundation-iasb-sign.jpg

The proposed amendments are as follows:

  • Relief from measuring and disclosing Scope 3 Category 15 GHG emissions associated with derivatives and some financial activities;
  • Relief from the use of the Global Industry Classification Standard, in some circumstances, in disclosing disaggregated financed emissions information;
  • Clarification on the jurisdictional relief to use a measurement method other than the Greenhouse Gas Protocol for measuring GHG emissions; and,
  • Permission to use jurisdiction-required Global Warming Potential values that are not from the latest Intergovernmental Panel on Climate Change.

The ISSB agreed to propose these amendments in January, following discussions of the Transition Implementation Group on IFRS 1 and IFRS 2 as well as the ISSB’s engagement activities. 

Entities can choose whether to apply the amendments’ reliefs, and jurisdictions can choose whether to adopt them without affecting the degree of their alignment with ISSB Standards. The reliefs would help preparers applying IFRS S2 by reducing the risk of duplicate reporting and the related costs associated with applying the standards.

“Proposing these amendments to a relatively new standard is not a decision that was taken lightly — we have carefully considered the need for such amendments and have sought to balance the needs of investors while considering cost-effectiveness for preparers,” Lloyd said. “Our due process is fundamentally important to us. We always consult our stakeholders when proposing changes to our standards and are balancing the need to respond to stakeholders’ needs on a timely basis with giving all interested parties the opportunity to participate in providing feedback by setting a 60-day comment period.”

Continue Reading

Accounting

House GOP drafts cuts to federal employee pension system

Published

on

Federal employee pension benefits are set to be pared back in Republicans’ giant tax and spending package working its way through the House, another slap at a workforce roiled by Elon Musk’s cost-cutting efforts. 

The proposal, which House Oversight Committee and Government Reform Committee Chairman James Comer announced on Friday night, would force many federal civilian employees to pay higher premiums for retirement benefits and to lower their eventual benefits by changing the formula for calculating payments. 

The Oversight Committee expects to vote this week on that plan and other workforce changes. If approved, they would be combined into legislation enshrining President Donald Trump’s legislative agenda, which Republicans aim to enact by August without the help of Democrats.

Comer said in a statement that the committee’s proposal would achieve “reduction in the federal deficit of over $50 billion.”

The biggest change would be to raise the premium that many long-time federal and postal employees pay out of their salaries in the Federal Employee Retirement System. Under the current system, contribution rates are arranged by the year an employee started: 0.8% in 2012 and earlier; 3.1% if hired in 2013, and 4.4% if hired in 2014 and afterward. The change would have employees pay 4.4% to raise $30.7 billion over a decade, according to the statement.

The proposal would also try to save $4.75 billion by basing retiree pension benefits on the highest five years of salary rather than the current three highest years. Those benefits are calculated based on top salary received and number of years in the U.S. workforce.

Other changes being proposed include eliminating supplemental retirement benefits for those who retire before age 62 and are unable to yet collect Social Security, and auditing family members of federal employees to see if they are eligible for health benefits.

Republicans argue that federal employee benefits are too generous compared to the private sector. At the same time, federal employees traditionally accept lower salaries than in the private sector partly because of the promise of those benefits. 

Republicans are planning a House floor vote next month on the larger legislation. The bill would then be sent to the Senate, where it could pass without the help of Democrats so long as the targets in the budget resolution the chambers already adopted are followed.

The House GOP has a goal of finding at least $2 trillion in savings to partially offset the cost of extending the 2017 Trump tax cuts, adding new cuts to taxes on overtime, tips and providing new breaks for older people and car buyers. The Senate has given itself leeway in its portion of the budget plan to make as little as $4 billion in cuts.

Continue Reading

Accounting

Art of Accounting: My first CPE presentation

Published

on

Complimentary Access Pill

Enjoy complimentary access to top ideas and insights — selected by our editors.

It’s hard for me to believe I published my Memoirs of a CPA book a year ago. Boy does tempus fugit. I was interviewed recently and asked how I got started presenting CPE programs, of which I’ve created and presented over 350 such programs.

I never thought about that and remembered the CPE program that got me started. It was around the mid-to-late 1970s and I had just joined the Managing of an Accounting Practice Committee of the New Jersey Society of CPAs. I was the youngest member of that committee. At the first meeting I attended, I pretty much sat there, listened and said nothing. Toward the end of the meeting the chairman mentioned that the speaker for the Managing Tax Season program that was two weeks away had to back out and they needed a speaker. He doubted they would get one this late. There was a little discussion and when it became evident they would have to cancel the program, I volunteered to present it. I said I never did a CPE program but would put together an extensive handout of forms and checklists I had developed in my practice. The committee members said to go ahead and I could call on any one of them for any help.

I have always been very process oriented and had my own forms, checklists and what has become my SmartProof worksheet (originally done by pencil on yellow worksheet paper). I made copies of all my checklists, forms, memos and sanitized copies of instructional letters to clients. I put together about a 30-page packet and referred to each page in my presentation. I had no notes but described every form and its use. 

What happened afterward was that I started getting calls with tax questions and got to know those colleagues. I was asked to repeat that program the following year and then started presenting it twice a year for the Society, in Northern and Southern New Jersey. Word got around and I was getting calls from other CPA societies and even the AICPA. Based on my AICPA presentation, I was asked to write articles in the Journal of Accountancy and then three books. My Managing Your Tax Season book for the AICPA had three editions. I’ve presented the original CPE program with additions each time I did it, probably 150 times with its many iterations. I still present it a few times a year. Evolving from this was my Reviewing Tax Returns book published by CPA Trendlines, which is currently in its third edition.

No matter who you are and whatever you’ve done, there is always a first time. My first CPE presentation was completely unplanned for and led to my becoming an “expert” in tax season processes. It also led to my developing my current 328-page 190-checklist Word file that I distribute for free to colleagues. If you want it, just email me at GoodiesFromEd@hotmail.com and put Checklists in the subject line. No messages necessary. Some years, over 10,000 checklist files are distributed for free.

I wholeheartedly recommend sharing your knowledge and experience by presenting CPE programs. It’s extremely satisfying, helps sharpen your knowledge, is great for networking with colleagues and sometimes there are very pleasant unintended consequences. If you don’t want to share what you know and do, imagine how it would have been for you if everyone else felt that way. If you want a prod or some help with ideas or getting started, contact me.

Do not hesitate to contact me at emendlowitz@withum.com with your practice management questions or about engagements you might not be able to perform.

Continue Reading

Trending